Hype Cycle for Human-Computer Interaction, 2014

Deep learning techniques are being developed to support near-real-time translation of dozens of languages and dialects. Gestural interfaces and mood recognition techniques are being applied in next-generation speech recognition to support full-sentence dialogue between humans and machines.

Despite these developments, there does not seem to be enough new transducer technologies or algorithm introductions to support the many other human-machine interface (HMI) concepts promulgating next-generation interfaces. These interfaces must be sophisticated enough to enable new concepts Gartner is introducing such as cognizant computing, which depicts context-aware machines able to make decisions on behalf of the user.

In fact, the average progress of technologies on the human-computer interaction Hype Cycle has drifted further away from the Plateau of Productivity in 2014, as older technologies reach the Plateau of Productivity and new interface technology concepts are explored. This phenomenon reflects a growing sentiment, among the human-machine interface technology community, that a breakthrough in human-machine interaction is imminent. However, this breakthrough will not be substantive without low-level technology innovation — a process that seems stalled at the moment.

Analysis

The 2014 human-computer interaction Hype Cycle reflects the wide variety of technologies being deployed to enhance the human-computer interface and their psychological and social ramifications. As underlying technologies evolve and costs fall, new opportunities emerge. Technology providers seek to address new markets by eliminating the barriers — between the user and the device — that have constrained use in the past for nontechnical users. To improve the odds of success, technology providers must do this without disrupting social norms.

This Hype Cycle consists of four profile types:

• Transducer — Technology related to interfacing with a human sensory system
• Algorithm — Technology for ascertaining human behaviors, characteristics and idiosyncrasies
• System — Hardware and/or software platforms that run some algorithms and require some type(s) of transducer(s)
• Concept — A notion, or idea, about human-computer interaction founded on one or a set of systems

The aggregate progress for concept profiles is expected to be lower than for system profiles, reflecting continually burgeoning ideas on how systems may be applied. However, the difference in progression between concept and system profiles should remain constant. A disequilibrium means that not enough new concepts are being generated or that systems are not maturing fast enough — as is the case in 2014.

Technology adopters should also consider the progress of the system profile as it relates to the transducer and algorithm profiles. A divergence would indicate an increase in adoption risk because maturing underlying technologies would not seem to be enabling systems.

Table 1 classifies all 51 technology profiles according to the four profile types, in order of decreasing progression toward the Plateau of Productivity. This relative positioning of each technology profile's position along the time axis (x-axis) of the Hype Cycle is a measure of progress. The aggregate progress of technology profiles across all four classes has decreased since 2013 (see Figure 1). New concepts are being introduced in the field of human-computer interaction with respect to human augmentation — namely, augmented-human staffing — which subsequently reduce the overall maturity level of the category.

Table 1. Technology Profile Classes in the 2014 Human-Computer Interaction Hype Cycle

Concept	System	Algorithm	Transducer
Rich Presence	Large Surface Computers	Speech Recognition	Bone Conduction
Virtual Reality	Wearable User Interfaces	Biometric Authentication Methods	Electronic Paper
Virtual Worlds	Heterogeneous Architectures	Handwriting Recognition	MEMS Gyroscopes
Augmented Reality	Ensemble Interactions	Gaze Control	Haptics
Machine Learning	UXPs	Gesture Control	3D LCDs
Affective Computing	Muscle-Computer Interface	Embedded GUI	Ambient and Glanceable Displays
Human Augmentation	Smart Robots	Facial Recognition	Pico Projectors
Quantified Self	Chip Implants for Nonmedical Applications	Natural-Language Question Answering	Head-Mounted Displays
Cognizant Computing	Tangible User Interfaces	Sensor Fusion	Smart Fabrics
Augmented-Human Staffing	Brain-Computer Interface	Mood Recognition	MEMS Displays
Virtual Personal Assistants	Olfactory Interfaces	Speech-to-Speech Translation	Flexible Display
	3D Video Telepresence	Behavioral or Gestural Analytics	Volumetric and Holographic Displays
		Emotion Detection/Recognition	Bioacoustic Sensing
		Gait Recognition
		Programming by Example (PbE)

Source: Gartner (August 2014)

The aggregate progress of system profiles has also dropped. This condition speaks to an alarming lack of practical use cases, particularly given the rising maturity of underlying hardware and software technologies (transducers and algorithms).

Figure 1. Progression of Technology Profile Class Type

Source: Gartner (August 2014)

The crux of human-machine interaction technologies is to minimize user (or operator) error and to enhance the user experience. In the 2013 human-computer interaction Hype Cycle, Gartner presented a theory that a major cycle is led by a concept that influences the generation of systems, consequently inciting algorithm and hardware innovations. Given this theory, it seems the field of human-computer interaction is due for more low-level technology innovations to support the evolution of systems and concepts. It seems the current crop of aging algorithms and transducers are not the ones that will catalyze the adoption of concepts that promise natural, effortless interaction between humans and machines. Gartner expects that more work with optical and ultrasound technologies, along with more refined algorithms, is required to achieve this quantum leap.

New Technologies
Four technology profiles were added (see Figure 2):

• Sensor Fusion
• Augmented Human-Staffing
• Cognizant Computing
• User Experience Platforms (UXPs)

Obsolete Before Plateau
No technology profiles have been made obsolete in 2014.

Renaming of Technologies
Virtual Assistants is now Virtual Personal Assistants.

Mobile Robots is now Smart Robots.

Reclassification of Technology Profiles
Olfactory Interfaces is classified as a system type instead of a transducer type.

Machine Learning is classified as a concept type instead of an algorithm type.

Figure 2. Hype Cycle for Human-Computer Interaction, 2014

Source: Gartner (August 2014)

Because human-computer interaction technologies usually represent an easier way to perform an existing task, rather than a fundamentally new capability, most of the technologies are clustered in the moderate benefit area of the Priority Matrix (see Figure 3). The number of technology profiles rated as highly beneficial or even transformational dropped in 2014 because location-aware technology and multitouch displays — both extremely impactful on society's relationship with machines — have reached the Plateau of Productivity and are now off the Hype Cycle.

Figure 3. Priority Matrix for Human-Computer Interaction, 2014

Source: Gartner (August 2014)

The following three technology profiles have reached the Plateau of Productivity and are now off the Hype Cycle:

• Location-Aware Technology
• Multitouch Displays
• E-Book Readers

Analysis By: Roberta Cozza

Definition: Bioacoustic sensing captures natural acoustic conduction properties in the human body using different sensing technologies. An example of this technology is Skinput, which allows the skin to be used as a finger surface. When a finger taps on the skin, the impact creates acoustic signals that are captured by a bioacoustic sensing device. Variations in bone density, size and the different filtering effects created by soft tissues and joints create distinct acoustic locations of signals, which are sensed, processed and classified by software.

Position and Adoption Speed Justification: This technology is being developed by researchers from Microsoft and the Human-Computer Interaction Institute of Carnegie Mellon University in Pittsburgh. In a prototype system, researchers focused on touch inputs on the arm and hand, and created an armband device for sensing. They evaluated different input locations, such as the fingertips and along the forearm.

The technology can also be integrated to augment the experience with a pico projector that projects dynamic graphical interfaces onto the hand or forearm. For example, a telephone keypad can be projected onto the palm of the hand, allowing real-time dialing without the use of a mobile phone.

Researchers have also developed a scrolling interface for projection onto the forearm. Users tap the top or bottom of the UI to scroll up or down, or go back one level in the UI hierarchy. Users can perform a simple pinching gesture with their thumb and fingers. Accuracy of 95.5% for five input locations on the whole arm has been demonstrated.

The technology is in the early stages of development, and future efforts will need to improve on the noninvasiveness of wearable bioacoustic sensor devices. Additionally, the disturbance from acoustic signals coming from other motions of the body will need to be reduced, particularly in walking or running scenarios (such as operating an MP3 player while jogging and using Skinput).

The input method is limited to quick skin taps, which in its current form does not permit more elaborate common gestures like sliding or dragging. Additionally, body mass index fluctuations can decrease sensing accuracy, and there is a high learning curve in setting up the solution.

Since its first appearance at Microsoft TechFest 2010, this project has remained under development, with no commercial product becoming available or expected in at least the next five years. Latest developments include a focus on a significantly decreased size of armband and improving accuracy of input. Research is ongoing but we have no reason to move this technology forward on this Hype Cycle.

Another example is the bioacoustic system from AT&T Labs. They have developed a prototype bioacoustic data transfer system that can send digital keys living as vibrations through the body (bones), enabling a door to open only to the unique acoustic "signature" of the homeowner. This works in conjunction with piezo sensors and a device such a smartphone — if the bioacoustic signature matches on both the device and the door knob, the door unlocks. Another application that AT&T is looking into is the exchange of data between people, where contact information is transferred via a handshake.

User Advice: Advances in this technology should be monitored and considered in scenarios where users can benefit from always-available and easily accessible input without direct access to the keypad of a device, such as a mobile phone or portable music player.

Business Impact: Using the human body as an input surface is an interesting concept for UIs. It could enable consumers to use larger and easily accessible additional input surface areas for interaction, compared with the small surface areas offered by the touchscreens on handsets.

Users could benefit by having large surfaces for input without needing to carry extra items. In addition, this type of input would allow accurate "eyes-free" touch interactions, because of our natural sense of body configuration (proprioception). Unlike other external input devices, most interactions could be performed without looking at the surface of a device. Experiments have also demonstrated a good level of accuracy in the input. Other external input approaches, such as smart fabrics or wearable computing, typically require an input device to be built into a piece of clothing, which is more complex.

Benefit Rating: High

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: AT&T; Microsoft

Analysis By: Tom Austin; Brian Manusama; Kenneth F. Brant

Definition: A virtual personal assistant (VPA) performs some of the functions of a human personal assistant. It observes its user's behavior, and builds and maintains data models, with which it draws inferences about people, content and contexts. It does so to predict its user's behavior and needs, build trust and, eventually, with permission, act autonomously on its user's behalf. It makes everyday tasks easier (by prioritizing emails, for example) and its user generally more effective (by highlighting the most important content and interactions).

Position and Adoption Speed Justification: VPAs represent a "perfect storm": a compelling vision, a great leap forward in technology, plentiful supply, and significant demand driven by transformational benefits.

Vision:

Apple's 1987 video "Knowledge Navigator" envisions a VPA.

The head of Microsoft's artificial intelligence (AI) research provides more recent examples in the video "Making Friends With Artificial Intelligence: Eric Horvitz at TEDxAustin."

Technology:

There are new and better algorithms (such as deep neural nets), much better hardware, and large bodies of information (big data) with which to train the systems underlying VPAs.

Supply:

There are already scores of VPA precursors, which lack one or more of the defining characteristics of VPAs. Precursors include virtual assistants in customer service applications (such as Nuance's Nina), conversational agents (such as Apple's Siri), and contextually aware proactive search features (such as those emerging in Google Now).

Google's Gmail Priority Inbox, introduced in 2010, is a narrow-scope VPA that organizes the user's email based on analysis of past behavior and content. Microsoft and IBM are expected to introduce similar capabilities by the end of 2014. Microsoft's new Outlook feature is code-named "Clutter," while IBM's first VPA will appear in "Mail Next."

We predict that Google, Microsoft and IBM will introduce more fully featured, opt-in VPAs in their cloud office systems in 2015 and 2016. At the Google I/O conference in 2013, Google outlined its "Knowledge Graph" efforts. At its SharePoint Conference 2014, Microsoft described its "Office Graph" and a client code-named "Oslo." Both look like strong precursors to more fully featured, conversational, opt-in VPAs that are likely to appear in the medium term. IBM has yet to reveal its plans, but we expect it have a lot to offer in the same time frame.

Venture capital investments in AI-related businesses are booming, and many startups are being acquired very early on, leading us to believe that there will be no shortage of supply of VPAs (or their subsystems and precursors).

Demand:

Initial demand for VPAs is likely to be driven by individual "bring your own" experiments, followed by more serious investigations by enterprises into whether VPAs can deliver a transformative advantage. Since the late 20th century, most progress in end-user-facing ad hoc tools has been disappointing, due to a lack of compelling new user benefits. VPAs may be the first new technology this century to present a real justification for investing ahead of everyone else.

This will not be a "winner takes all" segment. There will be many different VPAs for individuals and enterprises to consider. Individuals may use several VPAs with different specializations, such as health-related VPAs to help with diet, exercise, the quantified self, relationships and psychological wellbeing; VPAs to serve as personal shoppers; personal-career development and financial-management VPAs; and others for office-specific tasks like calendar management, email handling and external information monitoring.

User Advice: IT leaders should:

• Encourage experimentation, while creating opportunities for employees to share experiences and recommendations. Lead by doing.
• Prepare for mail-centered VPAs first, followed by a blossoming of the full range of capabilities envisioned in Apple's 1987 movie — and more.
• Recognize that privacy, security and innovation are at odds. Watch cautiously while encouraging experimentation. Imposing too many controls too soon due to a lack of trust in your employees could eliminate the opportunity to outflank competitors. Equally, though, granting your employees too much trust could be self-defeating, unless you keep careful watch.
• Carefully measure the impact of VPAs on people's behavior and performance. Use an ever-evolving set of metrics, identified by observation and crowdsourcing.

Business Impact: VPAs have the potential to transform the nature of work and the structure of the workplace. They could upset career structures and enhance workers' performance. But they have challenges to overcome beyond simply moving from research labs to product portfolios. It is far too early to determine whether, or how, they will overcome privacy concerns (although opt-in requirements make sense). Individuals will think long and hard about what they want each VPA to see and who else might view that information. Similarly, enterprises will be concerned about employees exposing confidential information via VPAs.

Benefit Rating: Transformational

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: Apple; Google; Highspot; IBM; Microsoft; Nuance

Analysis By: Jackie Fenn; Mark Raskino

Definition: Augmented-human staffing is the business capability of attracting, applying and risk managing the distinctive abilities of staff who elect to augment their physiology.

Position and Adoption Speed Justification: Human augmentation enhances cognitive and physical capabilities to achieve performance that goes beyond normal human limits. Augmentation examples include increased physical strength through exoskeletons, improved perception through sensory transference or adjustable aids (for example, a hearing aid with a phone app to optimize directionality), and enhanced concentration and learning through electrical brain stimulation. As technology to augment human capabilities makes the transition from mobile and wearable to implantable, organizations will need to confront a growing range of opportunities and issues relating to employees who choose — or in some cases are required — to enhance their bodies and minds though technology. For example, how will organizations need to modify their bring-your-own-device policies when the device is implanted in the employee?

Safety and privacy issues relating to augmented-human staffing are already surfacing, such as fears that wireless connectivity to pacemakers could be hacked and the fact that some states have pre-emptively outlawed companies from requiring an employee to have an RFID chip implanted.

At some point during the next decade, augmentation capabilities will become sufficiently safe and effective that a growing number of employees will start to electively "upgrade" themselves. There is evidence that workers are already enhancing their bodies with chemicals for work performance — for example, entrepreneurs using modafinil to increase productivity (see "The Real Limitless Drug Isn't Just for Lifehackers Anymore," New York Magazine, 31 March 2013) and Wall Street traders taking testosterone treatments to enhance aggression (see "Keep Taking the Testosterone," Financial Times, 9 February 2012). Quite serious cosmetic surgery has become normal for film actors and TV presenters over the past two decades, and it has spread into jobs where looks matter, such as the fitness industry.

Based on elective augmentation trends and the growing range of augmentation opportunities available, we are positioning augmented-human staffing past the Trigger of the Hype Cycle, even though it will be well over a decade before a significant number of organizations are affected. In the meantime, some organizations will contemplate offering their staff augmentation opportunities to increase performance, or creating policies to govern augmentation trends.

User Advice: Companies that require extreme talent to compete in their industries will have to confront the ethical and legal dilemmas of whether to employ people who have chosen to upgrade themselves. This might be in physically demanding task areas such as deep sea petroleum engineering, or in brain capability and concentration areas such as stock market trading. The competency of understanding where the boundaries are and what it takes to engage and manage people who are prepared to augment themselves will take human capital management into a new dimension: the intersection of technology, biology and ethics. New risk measurement, management and monitoring techniques will need to be created.

Business leaders should not turn a blind eye and allow this to start happening without policy, as it will be far more disruptive than earlier waves of "consumerization." Reputation and legal compensation risks will be unknown but high. Peer pressure issues within workgroups may damage morale. However, the individuals who start to improve themselves may do so covertly. CIOs and IT security professionals in companies likely to encounter the first augmented staff (most likely where high compensation is tied to correspondingly high performance, such as trading or high-stakes sales and deals) should therefore consider how to detect instances of augmentation — for example, by monitoring for anomalies in performance and achievements.

Business Impact: The highest business value will go to those companies that rely on small numbers of very high capability staff (for example, those in the top 5% of professional ability) and those with demanding physical requirements (for example, military, emergency services and sports). Business situations where additional staff numbers bring diminishing returns might also benefit. A business relevance test might be whether a 20% improvement in the professional mental or physical ability of critical staff would make a material difference to business results. Because augmented-human staffing as a deliberate strategy will be a difficult new capability to evolve, it is likely to be highly differentiating for the few enterprises that learn how to do it well (and how to address the ethical implications) in its first decade or two. Organizations will also need to establish the relative value of augmentation versus the use of robotic labor for certain tasks.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Recommended Reading:

"TDCS Guided Using fMRI Significantly Accelerates Learning to Identify Concealed Objects," Neuroimage, 2 January 2012

"Dick Cheney Feared Assassination Via Medical Device Hacking: 'I Was Aware of the Danger,'" ABC News, 19 October 2013

Analysis By: Jessica Ekholm; Brian Blau

Definition: Cognizant computing is the next step in the evolution of personal cloud. It uses big data and simple rule sets in order to increase personal and commercial information about a consumer through four stages: "Sync Me," "See Me," "Know Me" and "Be Me."

Position and Adoption Speed Justification: In the next few years, cognizant computing and smart machines will become two of the strongest forces in consumer and business IT. Any company in the business of providing a service, using apps or selling devices will be affected by cognizant computing in some way. Gartner predicts that, by 2016, OSs such as iOS, Android and Windows will no longer define a consumer's choice of smartphone. Cognizant computing also heralds the next evolution of the personal cloud, as consumers switch their focus away from devices to apps and the cloud (with negative implications for smartphone vendors such as Apple and Samsung). We predict that, by 2015, most of the largest companies in the world will be using cognizant computing to fundamentally change the way they interact with their customers.

By amalgamating and analyzing data in the cloud from many sources (including apps, smartphones and wearable devices), cognizant computing will provide contextual insights into how people behave: what they watch, do and buy, who they meet, and where these activities take place. This will help companies increase the lifetime value of their increasingly fickle customers, improve customer care, boost their sales channels, and make their customer relationships more personal and relevant. In essence, this new development will help companies innovate and create new business opportunities.

Cognizant computing has four stages:

1. Sync Me — Apps, content and information are made available across devices and shared contextually.
2. See Me — Data is continuously collected about users and their devices to gain an understanding of users' context.
3. Know Me — Understanding users' wants and needs, and proactively offering products and services based on pattern recognition and other machine-learning approaches.
4. Be Me — Developing intelligent apps and services that act on users' behalf.

At the moment, most activity is around the first two stages. As big data and the Internet of Things (IoT) become more pervasive, the vast amounts of information produced will enable complex systems to become more "intelligent," offering brand new opportunities in the latter two stages. This won't be without challenges or risk, however. Critical issues that will have to be addressed include consumer privacy, quality of execution and becoming a trusted vendor.

Cognizant computing is beginning to take shape via many mobile apps, smartphones and wearable devices that collect and sync information about users, their whereabouts and their social graph (mainly in the "Sync Me" and "See Me" stages of development). In addition, we are seeing the first personal digital assistants appearing with Microsoft Cortana, calendaring apps such as Tempo AI and, to an extent, Apple's Siri and Google Now. In the next two to five years, the IoT and big data will meet analytics, and more data will make systems smarter. By 2017, smartphones will handle some tasks for us better than we can do ourselves. At that point, consumers' personal clouds will interact with their smartphones and other devices, and the intricate app ecosystems they have created. The vast amount of very personal data that will flow between users and the brands is likely to be of some concern to users who do not necessarily want to share this amount of data. As this is a cause for concern for all involved, it would be advisable to work on how to create a trusted relationship between the user and the brand at an early stage of the relationship and, of course, to add strong privacy and security controls at all consumer touchpoints.

User Advice:

• Cognizant computing drives innovative analytics, apps, data and devices. Use its four-stage framework to stage new business models that can be used to identify supporting services opportunities, app and device features, and as a link to smart machines and the digitalization of business.
• Many cognizant computing services are still evolving, but look for companies that are developing deep learning, analytics, sophisticated algorithms and location-based technologies.
• Evaluate which cognizant computing assets you currently have and which you can either create in-house or be without; also consider which partnerships you need to go into to create a strong set of assets over the next 24 months, to reap future revenue.

Business Impact: We predict that most of the world's largest 200 companies will utilize the full toolkit of big data and analytical tools to refine their offers and improve their customer experience by 2015. Thus, over the coming two to five years we expect consumer-focused companies to use cognizant computing techniques to an increasing degree. This in turn will have big impact and affect the entire ecosystems and value chains across IT.

Technology and service providers that see opportunities within cognizant computing in the early years will create strong early-mover advantages, meaning they will be better equipped to develop stronger, more reliable ecosystems and reap early revenue benefits. They will also be able to deal with threats and issues with greater ease than late movers. At present, in 2014, no vendor has a full set of cognizant computing capabilities. That said, vendors such as Amazon, Apple, Facebook, Google and Microsoft have considerable collections of individual capabilities, while a host of smaller, more niche players such as Anki, Medio and Tempo AI have some interesting propositions. The big vendors are already racing ahead and consumer-focused businesses that have not yet got into this space are in danger of falling behind quickly.

Benefit Rating: Transformational

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Amazon; Apple; Facebook; Google; Here

Analysis By: Sylvain Fabre

Definition: Programming by example (PbE), also referred to as "programming by demonstration," is used to teach computers new behavior by demonstrating actions used on concrete examples — for example, sequences of operations that the human user may be going over repetitively in different situations.

Position and Adoption Speed Justification: At present, most PbE systems have been prototypes, widely inapplicable and generally confined to academic research.

User Advice: Although robot and computer-aided design learning applications are already being prototyped and should be monitored, it is too early for communications service providers to include PbE functionality in road maps for mobile networks and applications. However, potential uses of PbE in network performance management automation — for example, in self-organizing networks — should be monitored, although it appears that other avenues of machine learning are more promising in that area, such as anomaly detection.

Business Impact: Human-computer interfaces will be incrementally improved once PbE is leveraged by the computer — whether the user is demonstrating a task or simply performing it as usual.

PbE will influence user interfaces for Internet services and allow users who are not skilled programmers to effectively program computers and interfaces. Applications of PbE to computer learning environments for data analysis and statistics may be useful in big data applications, with instructible agents that gather and integrate heterogeneous data, for example. Basic, repetitive tasks for robotics systems could also be programmed using PbE, through user demonstrations.

The potential impact on mobile communications could be even greater, when context such as haptics, location and speed is added to the actions available on the user interface. Mobile devices could become highly personalized by users utilizing PbE approaches for the graphical user interface, OS or even mobility management.

In the area of mobile network infrastructure, self-organizing networks systems already seek to automate repetitive human actions, so PbE could become a promising area of improvement in addition to neural networks and artificial intelligence principles.

Benefit Rating: High

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Analysis By: Adib Carl Ghubril

Definition: 3D video telepresence facilitates collaboration by providing disparate parties with a lifelike visual experience, such that participants perceive themselves to be all together in one place.

Position and Adoption Speed Justification: Early and current commercial video telepresence installations, such as those from Cisco (acquired Tandberg), Polycom (acquired HP Halo), Avaya (acquired Radvision), Logitech LifeSize and Huawei Telepresence use flat, high-definition TV screens and specially lit rooms and require heavy consumption of bandwidth. These should be considered partial telepresence solutions because, ideally, telepresence attempts to implement all the human sensory elements remotely, but practical deployments are limited to sight, sound and manipulation (omitting smell and taste). Haptic teleoperations developments or telerobotics (for manipulation) have tended to be undertaken quite separately from telepresence. Telepresence is distinct from virtual presence, which gives participants the sensation of being immersed in a simulated virtual reality environment.

Companies such as DVE, Actis Technologies and TelePresence Tech are promoting some advances in video telepresence, including 3D stereoptic displays and eye-level optically embedded cameras, but these solutions are discounted here because they do not satisfy all three criteria for 3D telepresence, which are as follows:

• Creating a 3D image, either with a volumetric display or through stereopsis (each eye gets its own image)
• Rendering view-dependent images (changing what is viewed on the screen based on the position of the user's eyes relative to the projection)
• Focal accommodation (projection depth and focus)

Nevertheless, 3D telepresence is gaining momentum now that 3D camera technology is relatively stable and applied across a wide breadth of use cases. These systems, still found primarily in laboratories, will use a stereoscopic depth camera system to generate a 360-degree map of the source, which is then reconstructed at the destination with the stereoscopic projector's system or volumetric displays. The image is then superimposed on a background appropriately matched to the surroundings at the destination, such that attendees will feel like the display is an extension of their surroundings. Furthermore, because different portions of the 360-degree image will be shown to those attendees as they move about, they will feel like the people being telecast are physically present.

Microsoft's Applied Sciences Group has demonstrated a system, using a projector and waveguide, to display two sets of images — a left/right image for person 1 and a left/right image for person 2 — tracking the two people as they change their orientation relative to the display. For example, a participant who stands up and walks to one side of the display expects to, and indeed does, view the side of the telepresence participants. Microsoft is also planning to apply some of this technology to its Skype app.

Only a handful of companies have a working commercial setup for this, and these target conference applications and staged events. Though companies like DVE and Musion have some new installations, there don't seem to be notable new entrants. Furthermore, though some new products were announced during the past 12 months — such as DVE's telepresence stage, on which a PowerPoint presentation appears suspended in air, and the telepresence podium, where the speaker(s) appear in 3D — the systems are still restricted by slow innovation in volumetric displays.

User Advice: Although commercial telepresence systems that provide 3D images and eye-level contact among the participants are available, true 3D systems are still being developed and should be considered only as an R&D endeavor.

Meanwhile, consider enriching your stage presentations during company town hall meetings or industry conferences using sophisticated 3D telepresence systems that allow you to dynamically introduce animated characters, for example, and create special effects with presenters. 3D telepresence will also appeal to users in high-touch collaboration end applications, such as remote surgery, distance education or manipulation in hazardous environments.

Early practical 3D video telepresence deployments are likely to require special room and lighting formats, much like those needed to display Pepper's Ghost holograms (literally using smoke and mirrors to create the illusion) such as Musion's.

Business Impact: Business travel is a variable cost that affects all organizations, whether they are for-profit or not and regardless of industry or sector. The process of reaching an agreement between two or more parties, whether it concerns a purchasing decision, a course of action or a concept, involves cognitive and social factors. Cognitive factors, such as reasoning and the discussion of facts, can typically be adequately undertaken without all parties being in the same physical space. Social cues, however, are more difficult to convey in a teleconference, yet they are often critical to effectively "closing the deal."

If the old adage is true and "people do indeed buy from people," whether it is a product, service or idea, then the benefit of 3D telepresence is clear because it allows businesses to maximize the probability of a constructive interaction while minimizing the often-prohibitive travel costs associated with bringing people together.

Applications are likely to be led by entertainment (to offset high initial costs); enterprise executive communications (for high impact); distance learning (industrial training or in education); legal/judicial (avoids moving prisoners); medical (remote imagery); banking (boardroom, not customer use); high-end retail promotions; employment recruitment interviews and displacement of international travel.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: DVE; Microsoft; Musion; TelePresence Tech

Analysis By: Adib Carl Ghubril

Definition: Olfactory interfaces synthesize odorants from a digital signature or recognize odor by distinguishing the chemical properties of airborne particulates.

Position and Adoption Speed Justification: Olfactory interfaces have been utilized for commercial purposes, during the past decade, as an input modality — an electronic nose — and an output modality — an olfactory display.

Olfactory displays consist of a palette of odorants, a flow delivery system, and a control algorithm that determines the mixing ratios, concentration and timing of the stimulus. The palette stores odorants in either gel or microencapsulated form, and the diffuser is typically a set of nozzles equipped with microvalves or inkjet printer nozzles that release gels or liquid droplets by means of electrostatics or heat.

Electronic noses consist of an aroma delivery system that transfers volatile molecules to the transducer array, typically metal oxide or electrochemical, housed in a chamber in which air temperature and humidity are controlled and compensated for. Analog front-end and data converter circuitry then conditions and digitizes the transducer signal, whereby a statistical analysis of classification and recognition is performed.

Predominantly academic efforts have been commercially enhanced by the advent of the Digital Olfaction Society, which meets yearly; last year, it showcased innovations from Scentcom and Scentee. The Scentcom system mixes scents, dispensed from cartridges, into an air tunnel, which then infuses the user's near environment. This is done in synchrony with the content being consumed and in controlled bursts to maximize responsiveness. The Scentee system is simpler; it hooks into a smartphone, via USB, and essentially emits a scent as a means to provide notification or enhance a gaming experience. The canister contains only one type of scent, so the system can capture or discern only one event type at any one session.

Though these are promising advancements, olfactory interfaces must continue to improve. System size, reliability and weight are critical in today's mobile world, which relates to requirements for compressed air, an adequate odorant palette, mechanical reliability of the dispenser, and suitable storage for the gels and microcapsules. Furthermore, scent propagation should be such that it allows for multiple scents to be deployed without saturating the space or improperly mixing scents. This means the timing and intensity of injections should match the properties of the space and the content or message supported. Finally, the system should be self-calibrating, which means it needs to be able to adjust for changes in temperature, for example.

Complete interfaces would ideally also sense users' odors to help assess their mood, state of mind and health. These sensors must detect changes in aliphatic acids and other organic secretions induced by fluctuating hormonal levels brought on by stress, illness, ovulation or arousal, for example.

User Advice: Studies have shown that the addition of olfactory cues to visual and auditory ones makes an environment more memorable. Studies of visitor behavior at Disney World and Universal Studios, where olfactory effects have been used, attest to this. Scents have also been used in the enterprise to give workers a sense of urgency, among other things, and have been used in hospitals to help calm the patients. Though these findings are still at the research stage, consider these types of interfaces to raise productivity and improve your marketing positioning.

Electronic noses can detect the chemical markers associated with the onset of cancer or warn people of the onset of conditions such as yeast or sinus infections. They may also be used to detect halitosis.

Business Impact: Because scents evoke salient spatial cues, informing a person about being at a bakery, a farm or chemical plant, olfactory displays may be used not only to enhance the sense of presence in simulated environments, but also to train emergency personnel. Wearable sensors that warn emergency responders of high ammonia and hydrogen sulfide levels could save lives.

Olfactory displays are tentatively being used in retail settings to enrich the experience of shopping for cosmetics, perfume or liquor. Furthermore, taking cues about the state of mind and health of potential customers will help make for better decisions, more responsive machines and more compelling computer interaction, which will lead to stronger customer ties.

Noninvasive methods for detecting the onset of cancer can enable portable medical diagnostics and help curb healthcare costs.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: AromaJet; MicroFab; OlfaCom; Scentcom; Scentee; The eNose Co.

Analysis By: Mike Gotta; Whit Andrews; Frank Buytendijk

Definition: Quantified self is a movement promoting the use of self-monitoring through a wide variety of sensors and devices. Applications or services based on user data about activities, biometrics, environment and experiences provide a higher level of value from wearable and mobile devices, mobile apps, sensors and other "things" that offer self-tracking analytics, cross-sensor aggregation, social facilitation, observational learning and individualized coaching. Many different entities will provide these applications.

Position and Adoption Speed Justification: Analysis of this data allows individuals to gain a better understanding of their experiences and improve their wellbeing. Integration with social media allows users to connect with peers, share information, gain community support and learn from others. The quantified self movement has become a catalyst for the socialization of new types of technology and behavior. However, we now believe it will take five to 10 years before these are adopted by the mainstream due to cultural concerns (surveillance), societal acceptance (etiquettes), and business model fluctuations.

Although there are multiple types of applications, the most successful commercial implementations can be found in sports, fitness and health. There are thousands of fitness and health-related apps in smartphone app stores. Although application scenarios are broad, the dominant use case focuses on motion trackers and vital-sign monitoring (blood pressure and heart rate). However, application scenarios are expanding into areas such as mood monitoring and food/nutrition.

The breadth of devices itself is evolving rapidly as well. Many objects are being turned into sensor-based devices, including helmets, sneakers, glasses, watches, clothing and jewelry. The popularity of these devices and the immaturity of the technology can sometimes cause privacy, stability and quality issues. Proliferation of devices and apps without standards-based interoperability has created a market opportunity for new entrants to focus on data aggregation and normalization. Quantified self is also beginning to move into the workplace. For example, the inclusion of wearable devices and self-tracking apps as part of corporate wellness programs is becoming an aspect of employee engagement and digital workplace initiatives. Strategists are also looking at the potential of quantified self to improve personal and business productivity.

User Advice: The number and variety of personal devices and self-tracking mobile apps that collect data and provide feedback to users is increasing. Many different entities such as device makers, brands, software vendors, health-related firms, and developers of virtual personal assistants and smart machines will provide these applications.

While a dedicated community of people are interested in quantified self as a life philosophy to improve their own well-being, there are other populations interested in it to obtain medical insight or improve more serious health conditions — for themselves or in their caregiver role.

Marketers, innovation teams and community strategies should examine quantified self to help create a more social and collaborative brand experience, while leveraging personal analytics to establish greater customer intimacy.

Business Impact: Business strategists should ensure that proper policies and controls are in place to address user privacy concerns related to sharing personal data gathered via wearable devices, sensors and mobile apps. Organizations also need to invest in community management processes, and ensure that the personal participation needs and goals of community members are addressed. As people connect with peers, build relationships and interact with each other through the use of wearable devices, sensors and mobile apps, there may be a need for customized applications and unanticipated integration with other sites or internal systems. There are also behavioral, cultural and societal factors that come into play that strategists need to address early during design activities.

As more people use mobile and social technologies to collect and assemble data about themselves and their immediate surroundings, business opportunities emerge to apply insights gained from personal analytics and community participation to improve brand/customer relationships and product/service innovation. Within the workplace, organizations can create quantified self-incentives or requirements for employees to apply such analytics to measure performance or well-being, or to track employees in hazardous environments for health and safety reasons.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Fitbit; Jawbone; Nike

Analysis By: Adib Carl Ghubril

Definition: Gait recognition entails identifying humans and assessing their intent by matching the way they move and carry themselves with a database of known spatiotemporal profiles. Therefore, gait and posture recognition first entails detecting the human, then extracting material features and finally comparing those with a bank of known profiles.

Position and Adoption Speed Justification: While most of the work in this field is still academic, it is based on the same process used in the more mature technologies of facial recognition. Detecting a human starts by defining the background, subtracting it and subsequently generating a silhouette that is defined by height, torso, leg and stride vectors. Those vectors are then compared using statistical means to known models.

Although the analytical approach is still vision-based, the feature extraction algorithms are being refined to make use of texture information. Nonetheless, the work is still undertaken largely at research institutions. Gait recognition is also challenged by a changing background, which must be subtracted from the image to isolate the silhouette under study, and it is also challenged by the silhouette changing direction or changing activity.

Gait recognition research is also being used to ascertain suspicious behavior by helping distinguish between a loaded gait and a normal gait. For example, a pedestrian exhibiting loaded gait patterns might be an indication that the content of the backpack being carried is inordinately heavy and therefore suspicious.

Refinements continue in research institutions, which raise hype a little. However, no commercialized prototype has yet been perceived, so the technology remains embryonic.

User Advice: Gait recognition offers the means to perform unobtrusive biometric assessments because it is idiosyncratic and can be performed using vision-based analytics. However, it is also subject to similar limitations from which facial recognition systems suffer — namely, occlusions and lighting conditions. Nevertheless, security professionals should consider how the application of gait recognition could improve the robustness of surveillance systems.

Furthermore, gait recognition can help track the activity of patrons in a retail environment or in shared spaces, such as airports. This would allow administrators to not only better understand how behavioral trends are affected by surroundings, but also better manage the flow of pedestrian, traveler or shopper traffic.

Business Impact: Improving the accuracy of tracking systems can help architects design better and safer public spaces. It also can literally make a difference between life and death to unwitting victims because the surveillance system can match the activity patterns it detects with a library of activity profiles that are deemed suspicious, thus sounding an alarm. Gait recognition is an invaluable innovation, but because it works in conjunction with other technologies such as facial recognition, it may not stand alone as a transformational technology.

Benefit Rating: High

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Analysis By: Jackie Fenn

Definition: A brain-computer interface is a type of user interface, whereby the user voluntarily generates distinct brain patterns that are interpreted by the computer as commands to control an application or device. The best results are achieved by implanting electrodes into the brain to pick up signals. Noninvasive techniques are available commercially that use a cap or headband to detect the signals through external electrodes.

Position and Adoption Speed Justification: Brain-computer interfaces remain at an embryonic level of maturity, although we continue to advance them slightly along the Hype Cycle to acknowledge the growing visibility of several game-oriented products (such as those from Emotiv and NeuroSky) in the emerging field of neurogaming. The major challenge for this technology is obtaining a sufficient number of distinctly different brain patterns to perform a range of commands — typically, fewer than five patterns can be distinguished. However, this proves sufficient to play interactive games and control equipment or even some vehicles. One approach that operates well within these constraints is to watch for the distinctive brain pattern associated with recognizing a desired goal — for example, brain-driven typing flashes letters on the screen until the desired letter is recognized by the user's brain. Further advances are likely to arise from research on activating prosthetic limbs, whereby functional magnetic resonance imaging (fMRI) and other brain-scanning techniques are being used to identify people's natural brain patterns when performing various actions (such as closing their hands). fMRI is also proving effective in reading emotions and determining what type of object a person is looking at or thinking about. The Obama administration's decade-long Brain Activity Map project will also drive improved interpretation of brain signals. Several of the commercial systems also recognize facial expressions and eye movements as additional input.

Outside of medical uses, such as communication for people with "locked in" syndrome (a condition in which a patient is aware and awake but cannot move or communicate verbally), other hands-free approaches, such as speech recognition, gaze tracking or muscle-computer interfaces, offer faster and more-flexible interaction than brain-computer interfaces. The need to wear a headband to recognize the signals is also a serious limitation in most consumer or business contexts. Researchers at Brown University have succeeded in reading brain signals from a low-power wireless system implanted in animals for more than a year, paving the way for research on human brain signal implants during the next decade.

User Advice: Treat brain-computer interfaces as a research activity. Some niche gaming and disability-assistance use cases might become commercially viable for simple controls; however, these will not have capabilities that will generate significant uses in the mainstream of business IT.

Business Impact: Most research is focused on providing severely disabled individuals with the ability to control their surroundings. Commercialization is centered on novelty game interfaces and applications that help users become more aware of their own brain state, and thus, they are better able to relax or focus. As wearable technology becomes more commonplace, applications will benefit from hybrid techniques that combine brain, gaze and muscle tracking to offer hands-free interaction.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: Brain Actuated Technologies; Emotiv; InteraXon; neurowear; Neural Signals; NeuroSky; Personal Neuro Devices

Recommended Reading:

D. Orenstein, "Brown Unveils Novel Wireless Brain Sensor," Brown University, 28 February 2013

L.R. Hochberg, D. Bacher, B. Jarosiewicz, N.Y. Masse, J.D. Simeral, J. Vogel, S. Haddadin, J. Liu, S.S. Cash, P. van der Smagt, J.P. Donoghue, "Reach and Grasp by People With Tetraplegia Using a Neurally Controlled Robotic Arm," National Center for Biotechnology Information (NCBI), 16 May 2012

Analysis By: Jackie Fenn

Definition: The field of human augmentation focuses on creating cognitive and physical improvements as an integral part of the human body. An example is using active control systems to create limb prosthetics with characteristics that can exceed the highest natural human performance.

Position and Adoption Speed Justification: Human augmentation moves the world of medicine, wearable devices and implants from techniques to restore normal levels of performance and health (such as cochlear implants and eye laser surgery) to techniques that take people beyond levels of human performance currently perceived as "normal." In the broadest sense, technology has long offered the ability for superhuman performance — from night-vision glasses (or even a simple flashlight) that help people see in the dark to a financial workstation that lets a trader make split-second decisions about highly complex data.

Although most techniques and devices are developed to assist people with impaired function, development of superhuman capabilities has started. Power-assisted exoskeletons provide increased strength and endurance to soldiers and caregivers. Hearing aids, such as the GN ReSound LiNX, offer their wearers superior hearing ability through wireless real-time adjustments on a mobile phone app; for example, these may be used to mute music and increase directional focus in a noisy environment. Researchers are experimenting with creating additional senses for humans, such as the ability to sense a magnetic field to develop the homing instinct of birds and marine mammals; and with sensory substitution, such as allowing a blind person to drive a car by translating visual information into vibrations. Brain stimulation techniques, such as transcranial direct current stimulation, are proving effective in enhancing concentration and accuracy. To date, these systems are worn or strapped onto the body, rather than surgically attached or implanted; but with advances such as thought activation of mechanical limbs, the distinction between "native" versus augmented capabilities will start to blur.

Increasing specialization and job competition are demanding levels of performance that will drive more people to experiment with enhancing themselves. Augmentation that reliably delivers moderately improved human capabilities will become a multibillion-dollar market during the next quarter century. However, the radical nature of the trend will limit it to a small segment of the population for most of that period. The rate of adoption will vary according to the means of delivering the augmentation. Drugs are already used extensively for off-label performance enhancement, such as anabolic steroids for strength and modafinil for alertness and concentration. Wearable devices are likely to be adopted more rapidly than those involving surgery, although individuals are already experimenting with implanting technology for purposes such as storage and listening to music. The huge popularity of cosmetic surgery is an indicator that even surgery is not a long-term barrier, given the right motivation.

Ethical controversies regarding human augmentation will emerge even before the technology becomes commonplace. Several states have already passed bills banning employers from requiring chip implants as a condition of employment. Future legislation will need to tackle topics such as whether an employer is allowed to prefer a candidate with augmented capabilities over a "natural" one. Longer term, the potential for genetic and epigenetic manipulation to improve desirable characteristics will further inflame deep ethical divides.

User Advice: Organizations aiming to be very early adopters of technology, particularly those whose employees are engaged in physically demanding work, should track lab advances in areas such as strength, endurance or sensory enhancement. Employers will need to weigh the value of human augmentation against the growing capabilities of robot workers, particularly as robots may involve fewer ethical and legal minefields than augmentation. Cognitive enhancement through technology is already represented by the growing use of — and dependence on — instant mobile access to information and community, and organizations must continue to be ready for consumer- and employee-led adoption of the latest wearable or even implantable technology. Organizations can gain an early understanding of some of the opportunities and issues by tracking the Quantified Self movement, which promotes self-monitoring through a wide variety of sensors and devices with a goal of improving physical and mental well-being.

Business Impact: The impact of human augmentation — and the ethical and legal controversies surrounding it — will first be felt in industries and endeavors demanding extreme performance, such as the military, emergency services and sports. In parallel, consumer applications using sensory enhancement through augmented reality (for example, collision alerts or "friend nearby" notifications) will be delivered initially through mobile or wearable devices.

Benefit Rating: Transformational

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: Cyberdyne; Raytheon

Recommended Reading:

"Conjuring Images of a Bionic Future"

"Blind Man Drives High-Tech Car at Daytona Speedway"

feelSpace belt for directional awareness

Analysis By: Stephen Prentice

Definition: Volumetric displays create visual representations of objects in three dimensions, with a 360-degree spherical viewing angle in which the image changes as the viewer moves around. Unlike most 3D planar displays, which create the illusion of depth through visual techniques (stereoscopic or autostereoscopic), volumetric displays create lifelike images in three-dimensional space.

Holographic displays can recreate a 3D image, but they are not true volumetric displays.

Position and Adoption Speed Justification: Volumetric displays have barely emerged from the laboratory. The iconic volumetric image of Princess Leia created by R2-D2 in the first Star Wars movie (released in 1977) remains an elusive, yet aspirational, goal.

True volumetric displays fall into two categories: swept volume displays, and static volume displays. Swept volume displays use the persistence of human vision to recreate volumetric images from rapidly projected 2D "slices." One approach is to project images onto a rapidly rotating mirror inside a protective enclosure (to protect viewers from injury, should they attempt to touch the images). Static volume displays use no major moving parts within the image display volume, but rather, rely on a 3D volume of active elements (volumetric picture elements, or voxels) that change color (or transparency) to create a 3D image within the display volume. Low-resolution displays may use transparent elements such as LEDs, while some higher-resolution displays use techniques such as pulsed lasers that are directed by scanning mirrors to create balls of glowing plasma at the location of each voxel.

Swept and static volumetric displays suffer from the significant dangers of rapidly moving parts or ionized particles in the vicinity of people, especially because the volumetric nature of the generated image convinces the brain that it is solid and "real" and, therefore, can be touched. In all cases, the volume of data required to generate a volumetric image is considerable — typically on the order of 1,000 times more to create a 24-bit voxel image (1,024 layers on the z-axis) than the corresponding 2D image. In all cases, the amount of CPU processing required is equally significant, compared with creating a 2D image.

Holograms can be deployed as an alternative to a volumetric display, but with a more restricted viewing angle. It should be noted that the term "holographic display" is frequently (but incorrectly) applied to any image that creates an appearance of 3D. Some current theatrical and conferencing displays allow realistic images to appear out of thin air and can, with care, allow individuals to walk "around" them. However, they are simply 21st-century implementations of the 19th-century Pepper's ghost illusion, using high-intensity projectors and Mylar display films, and not true volumetric or holographic displays.

Several companies, including InnoVision Labs, Sony and Realfiction, have demonstrated 3D or holographic images generated from their projectors, but not one of the images has been commercialized yet.

Competing with volumetric and holographic displays, 3D displays such as those increasingly found in televisions create a visual impression of depth, but rely on spatially multiplexed images that deliver different views to each eye and allow the brain to reconstruct a 3D representation. They are planar displays that simulate depth through visual effects, rather than true volumetric displays that create an image in a display volume with real depth.

User Advice: Outside of specialized areas, where budgets are not significant constraints, this technology remains firmly in the lab, rather than in commercial applications. Current technologies limit the size of volumetric space that can be displayed, and the mechanical solutions create potentially dangerous, rapidly moving parts. Until alternative approaches can be delivered (which seems unlikely in the near future), volumetric displays will remain an extremely niche product. Concurrently, the rapid growth and continuing development of 3D televisions in the mainstream markets threaten to overwhelm the continuing development of volumetric and holographic displays outside of specialized markets.

Business Impact: General applications are not well-developed for business use. To date, simple applications in marketing have been deployed — usually targeted at high-end retail environments, and there are some specialized applications for geospatial imaging to enhance 2D maps, and for use in architectural rendering. However, most of these can be achieved at much lower costs using other more-commercialized technologies, such as 3D displays. Potential application areas include medical imaging, consumer entertainment and gaming, and design, but costs will need to fall dramatically for these to be viable for using true volumetric displays.

Benefit Rating: Low

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: HP; Musion; Realfiction; Sony

Analysis By: Jackie Fenn

Definition: A semiconductor chip is implanted into a person and read by an external scanner. To date, nonmedical applications have focused on identification — similar to the technology widely used for pet identification.

Position and Adoption Speed Justification: Chip implants for medical purposes are developing along multiple fronts, including wireless connectivity for motor control (for example, Brown University's wireless brain sensor) and brain stimulation or inhibition to moderate food intake. Outside of medical applications, the primary use of chip implants has been for identification — such as for Alzheimer's patients (Florida), physical access to secure areas (Mexico) and private club access (Spain). However, there has been little activity on this front since 2010, when PositiveID, the only company selling chip implants approved by the U.S. Food and Drug Administration for human use, stopped marketing identification chips to concentrate on chips for medical applications. The other main area of activity is "body modding" through subdermal implants. This type of implant is focused primarily on esthetics (similar to a 3D tattoo), but some enthusiasts have implanted technology, such as cameras and magnets, as well as ID chips.

Renewed interest in chips for identification may be triggered by technology advances, such as increased storage, which allows the chip to store an entire medical record, or an active chip used for location tracking (although active chips raise the additional challenge of powering the chip by an implanted battery or by energy harvested from the body.) Broader adoption may also grow out of medical technology as implanted devices begin to enhance human capabilities (for example, vision and hearing), rather than just repairing it, and as able-bodied individuals begin to use the technology electively.

Health concerns and privacy issues have already surfaced, such as the reports of implanted chips leading to tumors in animals and the laws passed in several U.S. states making it illegal to require an individual to have a chip implanted. The inconvenience of upgrading the chip as the technology evolves is another barrier. Applications will need to show extremely high value to overcome these concerns, because current applications could be achieved by other, less-intrusive approaches — in particular, wearable devices. Emerging techniques, such as epidermal electronics, whereby a circuit is printed or stamped onto the skin, will also offer competition. Because identification is unlikely to be the application that takes this technology mainstream, and other applications are still experimental, we expect nonmedical chip implants to remain stalled on the Hype Cycle for several years.

The benefit rating of low is based on passive identification applications. More advanced uses in human augmentation could result in transformational benefits.

User Advice: There is little motivation to use chip implants purely for identification, because biometric identification offers most of the same benefits (such as being unique to each individual and obviating the need to carry a card). Other applications such as location tracking and implanted computer-brain interfaces for fully abled individuals remain experimental, with most current applications using wearable technology (for example, health monitoring bracelets) as a less-intrusive and more acceptable alternative.

Business Impact: The primary business impact of chip implants will remain in the medical domain, although leading-edge applications in brain-computer interfaces and human augmentation may benefit from this technology over the next few years.

Benefit Rating: Low

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Analysis By: Kenneth F. Brant

Definition: Smart robots are literally smart machines that have a physical form factor — unlike virtual personal assistants and smart advisors — and that can work autonomously in the physical world and learn from their experiences. Smart robots sense conditions in their local environments, recognize and solve basic problems, and learn how to improve. Some have a functional form, such as warehouse robots from Amazon's Kiva subsidiary, while others have humanoid appearances, such as Baxter from Rethink Robotics. They may work alongside humans or replace human labor.

Position and Adoption Speed Justification: While industrial robots have been around for a long time and are certainly more advanced in their life cycles, the subset of smart robots is much newer and has had significantly less adoption to-date. That is why smart robots are positioned at the midpoint between the Technology Trigger and the Peak of Inflated Expectations. Hype and expectations will continue to build around these smart robots over the next few years as a dynamic set of large and small suppliers develops more solutions across the wide spectrum of generic and industry-specific use cases. Several recent key events have expedited the adoption speed we now expect to see in this category: (1) the acquisition of Kiva Systems by Amazon and Amazon's subsequent plans to deploy 10,000 Kiva robots to fill customer orders by the end of 2014; (2) Google's acquisition of Boston Dynamics and seven other robotics companies within a six-month span in the second half of 2013 and its ability to incorporate machine learning in these acquired robot assets; (3) Rethink Robotics' launch of Baxter, which can work alongside human employees to perform simple assembly line tasks by being shown what to do (rather than requiring programming), at prices starting around $25,000; and (4) the transfer of military technology to commercial and consumer robotics from companies like iRobot. These events will create a competitive race on the supply side of the market to build scale in this category, now that we have witnessed initial pilots and limited trials on the demand side of the market. Users, too, will race to find competitive advantage after leaders in their industry segment have begun their journey with smart robots.

User Advice: Consider smart robots as real substitutes and complements to your human workforce. Begin pilots designed to assess product capability and maturity, feasibility to your business processes, and potential returns on investment. Users in e-commerce, manufacturing, distribution and retail segments, healthcare, and government services that have high labor costs associated with repetitive workflows and requirements for agility and worker safety should explore the benefits of smart robots in their operations now and plan their adoption in stages.

Business Impact: Smart robots will make their first business impact across a spectrum of product and service-centric industries. Their ability to do physical work, with greater potential reliability, lower costs, greater safety and higher productivity, is common across these industries. Their initial impact will be greatest in industries that have the highest cost of labor in commercial industry operations (like material handling and logistics in manufacturing, healthcare and retail) or face the highest risk to their human workforce in public-sector operations (like inspecting and defusing bombs or investigating natural disasters or other threats to citizens and military personnel). Their business impact will be in improving productivity, reducing the costs of labor in industries that face great employee attrition and training expenses, and greater agility in industries that have repetitive but changing work routines. The ability for organizations to assist, replace or redeploy their human workers in more value-adding activities creates potentially high — but not transformational — benefits. Typical and potential use cases include medical materials handling; hazardous waste materials disposal; prescription filling and delivery; patient care; direct materials handling; stock replenishment; product assembly; finished goods movements; product pick and pack; e-commerce order fulfillment; package delivery; shopping assistance and customer care; and citizen protection.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Aethon; Amazon (Kiva); Google; Honda; iRobot; Intelligent Hospital Systems; InTouch Health; Panasonic; Rethink Robotics; Swisslog; Symbotic; VGo Communications

Analysis By: Adib Carl Ghubril

Definition: In a tangible user interface (TUI), the user controls digital information and processes by manipulating physical, real-world objects (rather than digital, on-screen representations) that are meaningful in the context of the interaction.

Position and Adoption Speed Justification: Although TUIs will likely take most of the next decade to develop commercially because of the need to fundamentally rethink the user interface experience and application development toolset, we believe they provide the kind of tactile feedback that enriches the user experience. We therefore expect this technology to continue to evolve, particularly in applications where the user is crafting something.

"Thumbles," a product created by system-design house Patten Studio, combines three modes of interaction that allows users to see, hear and manipulate an application running on a surface computer. The manipulation is via miniature robots that are mechanically controlled by both the application and the user.

Tangible Play, a startup, introduced Osmo, an interactive puzzle and gaming application where collaborators must react to whatever they see on the screen with whatever objects they have at their disposal — puzzle pieces, block letters, and even pencil/paper. A light guide and stand allow Osmo to make use of the processing resources and sensors available on the tablet such that the screen, front-facing camera, and a rigid resting surface make up the gaming environment.

Meanwhile, Sifteo has made continuous improvements to its hardware over the past two years, and now offers a developer's software development kit.

TUIs have also been impacting music-creating applications of late. Reactable Systems allows users to superimpose musical beats and tones by varying the orientation and relative positioning of physical cubes on a surface. The cubes vary in size and are marked with unique symbols.

User Advice: TUIs can be a powerful educational tool for those experiencing the world for the first time. By handling a physical model of an animal, plant or object, infants can generate narrative, musical and visual feedback about the model, which stimulates their intellect in a more meaningful way.

Similarly, organizations with customer-facing, branded kiosk-style applications (such as retail and hospitality) should evaluate opportunities for a highly engaging (and relatively high-cost) customer experience using TUIs. Others should wait for new classes of devices and peripherals to integrate the capability seamlessly. Surface, or table top, computer makers could provide the catalyst for application developers to address use cases and raise adoption further.

Business Impact: TUIs will provide an intuitive way for people to bridge the physical and digital worlds. Applications focused primarily on customer-facing experiences, collaborative planning, and the design of real-world objects and places will benefit the most.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: Microsoft; Patten Studio; Reactable Systems; Samsung; Sifteo; tangible play

Analysis By: Jan-Martin Lowendahl

Definition: Affective computing technologies sense the emotional state of a user (via sensors, microphone, cameras and/or software logic) and respond by performing specific, predefined product/service features, such as changing a quiz or recommending a set of videos to fit the mood of the learner. Affective computing tries to address one of the major drawbacks of online learning versus classroom learning — the teacher's capability to immediately adapt the pedagogical situation to the emotional state of the student in the classroom.

Position and Adoption Speed Justification: True affective computing technology, with multiple sensor input, is still mainly at the proof-of-concept stage in education, but it is gaining more interest as online learning expands and seeks means to scale with retained or increased quality. A major hindrance in its uptake is the lack of consumerization of the needed hardware and software involved. It has to be inexpensively available for students because they use their personal devices before education institutions can deploy affective computing software. However, products such as Affectiva's Affdex or ThirdSight's EmoVision are promising because they enable relatively low-cost, packaged access to affective computing functionality, even if these particular products are geared toward testing media/advertising impact on consumers. Another industry, the automotive industry, is more advanced. Here, the technology has not yet found its way into mainstream vehicle production, but lightweight emotion detection — for example, being tired behind the wheel — is an option in trucks on the market today. Addressing issues such as driver distraction and driving while tired creates more awareness for mood sensing in a practical and ubiquitous product — the car.

The leading research lab in this field is MIT's Affective Computing Research Group, which has many projects and is working on sensors, such as wristband electrodermal activity sensors connected by Bluetooth to a smartphone, and software, such as the MIT Mood Meter, that assess the mood on campus based on frequency of smiles as captured by ordinary webcams. Developments like these can speed up the application of affective computing in education, but the road ahead still seems long due to complexity. It is possible that there needs to be a breakthrough in a more consumer-oriented area such as gaming before affective computing can be applied at a larger scale. One thing that might jump-start implementation would be if facial recognition services for identification and proctoring in online learning, from companies such as Smowl and KeyLemon, were implemented more often and if affective computing were sold as an add-on to that kind of service. An interesting and more specialized branch of affective computing involves robots such as the emote project. This "artificial tutor" approach has many interesting possibilities. It uses a robot's movements to strengthen affective feedback with the student, but it has the drawback of needing a physical robot. The latter is likely to make this approach more costly for education institutions and delay implementation.

Successful affective computing will most likely involve a complex architecture in order to combine sensor input and provide an accurate response in real time. Mobile learning via cloud services and handheld devices, such as smartphones and tablets, is likely to play a key role in the first few generations, with a larger market penetration due to the relatively controlled ecosystem it provides (high-capacity computing combined with a discrete device with many sensors). As content (for example, textbooks) becomes more digitized and is consumed on devices that have several additional sensors (for example, tablets with cameras and accelerometers), interesting opportunities will arise to mash up the capabilities of, for example, Knewton's Adaptive Learning Platform and ThirdSight's EmoVision, making affective computing for untutored learning more accessible. This could potentially increase the number of data points available for statistically based adaptive learning.

Altogether, this merits a position that is still in the trigger phase, with at least 10 years until it reaches the Plateau of Productivity.

User Advice: Most institutions should only continue to follow the research and development of affective computing in education and other industries. However, in order to be prepared for the strategic tipping point of implementation, institutions should start estimating the potential impact in terms of possible pedagogical gains and financial impact, such as increased retention for online learning. Institutions with a large online presence, or that want to exploit the hype for brand recognition, should get involved now. Partner with automotive suppliers, consumer electronics companies and universities (particularly online) to further explore this field.

Affective computing can involve collecting sensitive data about students, which makes it important to make sure that any privacy laws and concerns of the users are met (such as policy about if, when and how data is stored). Preferably, any use of affective computing should involve an "opt-in" process.

Business Impact: Affective computing is an exciting area with the potential to bring back a bit of the lost pedagogical aspects of classroom learning and to increase the personalization of online learning. One important advantage of this technology is that, even if it is inferior to a face-to-face student-teacher interaction, it scales well beyond the 100-plus-student lectures that today offer limited individual pedagogical adaptivity. A potential complement or competition to remedy the scalability problem is the social-media-based peer-mentoring approach, as exemplified by Livemocha and, more lately, by massive open online courses (MOOCs). In the Livemocha example, a sufficient scale of the community of quality subject matter mentors can be reached by tapping the full Internet community of more than 2 billion users.

In general, affective computing is part of a larger set of approaches to further personalize the educational experience online. Another example is adaptive learning that depends on the statistical data of learners in the same pedagogical situation. It is also related to context-aware computing in general.

The ultimate aim of affective computing in education is to enhance the learning experience of the student, which should result in tangible results like higher grades, faster throughput and higher retention. These results will benefit students, institutions and society.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: Affectiva; Affective Media; IBM; Pearson Education; ThirdSight

Analysis By: Anshul Gupta

Definition: Flexible displays are flat computing screen panels constructed of thin or flexible substrate that can be bent, rolled, folded or flexed without loss of functionality. The flexible substrate used to replace conventional glass substrate can be plastic or thin glass. The displays can be printed or deposited into thin foil.

Position and Adoption Speed Justification: Flexible displays include many components and supporting technologies, such as flexible substrates, conducting transparent conducting oxides and polymers, electro-optic and reflecting materials, inorganic and organic electronics and packaging technologies. It also requires new manufacturing processes, such as roll-to-roll manufacturing and coating and printing technology. There is no standard flexible display with standard component technology yet, as the technology is still evolving at both the component and display levels. Carbon nanotubes, graphene or ClearOhm ink are some of the conductive coating materials replacing conventional indium tin oxide to make flexible display a reality.

Developments in many components and technologies supporting the manufacturing of flexible displays have accelerated the concept of flexible display toward reality. LG introduced G Flex, a 6-inch curved phone with a flexible display. Samsung launched the Galaxy Round smartphone with a curved display that did not have flexible display. Samsung launched curved HDTV in 2Q14 and is working on flexible active-matrix organic light-emitting diode (AMOLED) displays with a film-like polyimide (PI) substrate that should bring flexible display-equipped Samsung smartphones to market in 2015. Plastic Logic has developed flexible tablets by printing organic thin-film transistors on flexible backplanes. Toppan is also using organic thin-film transistors for flexible displays.

Flexible display technology can result in many compelling applications, such as:

• Large wall-sized reflective screens for use in conference room settings that could be rolled away when not in use; however, it could be very costly.
• Small portable rollable displays.
• Irregular-shaped displays used in the steering wheel of an automobile.
• Conformed displays integrated in an automobile that fill up the entire dashboard.
• Wristband displays that are permanently conformed throughout their lifetime.
• Mobile phones, watches or tablets with flexible displays.

Ongoing innovation in flexible substrate, costly manufacturing processes, yield and evolving technology will keep costs high. Ink and material costs will play a big role in the commercialization of displays. Flexible displays may find a subset within glass-based displays for many applications, but it will be difficult for flexible displays to compete solely on cost alone in the inexpensive and small display module market or in the high-end, high-performance market (such as desktop and laptop screens).

User Advice: Flexible displays are becoming a reality, but complete flexible devices will still take more time. Consumer devices with flexible or curved displays are already available; a complete foldable smartphone, however, is still a few years away.

Expect devices equipped with flexible displays to have a high initial cost as the component technology is still evolving. Flexible display devices that serve an unmet need or bring new use cases for existing types of devices will justify investment in the near term. The mass displacement of existing types of devices, such as smartphones, tablets and notebooks, with flexible displays will be delayed over the long term until the technology matures.

Business Impact: Consumers generally prefer having just one device rather than owning multiple devices; a flexible display can meet user demand for a thin and light device. Flexible displays could bring together laptops, tablets and smartphones into a single device. You could unroll or unfold the device to the size needed based on what you need to do. For instance, a device can be unfolded to three inches by five inches to make a phone call, be increased to a 10-inch display to form a tablet, or open up all the way to a full 15 inches for a laptop and typing surface.

Flexible displays could be extremely disruptive to the display industry, bringing new form factors or completely disrupting current display technology ecosystems.

Benefit Rating: High

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: Cambrios; LG Display; Panasonic; Samsung Electronics; Sony

Analysis By: Alexander Linden

Definition: Machine learning is a discipline that allows software components to be synthesized from data without being explicitly programmed. The two major subdisciplines are: supervised learning, where classifications and predictions need to be performed (such as predictive analytics); and unsupervised learning, where clustering, dimensionality reduction or density estimation is the objective. Supervised learning is making rapid progress, with ensemble techniques and deep learning.

Position and Adoption Speed Justification: Gartner has repositioned machine learning. Even though as a discipline it is three to four decades old, it has recently regained enormous interest by becoming the major catalyst for a number of innovations (for example, Google's driverless car and IBM Watson's question-answer system), and its applicability is being significantly expanded by a surge in available data and business complexity. As a general concept it is already widely adopted via predictive analytics solutions (such as database marketing and cross-selling). Gartner anticipates, that new and important use-case scenarios and also increasing commercialization will keep it rising to even higher levels of attention.

User Advice: As the number of sensors and resulting datasets surge, machine learning must be considered as an alternative approach to traditional software engineering. Most data-intensive industries would be well advised to develop deeper in-house skills, as well as to cater to networks of local and international experts relevant to their ecosystem of business processes.

Business Impact: Machine learning is especially critical in situations where software engineering has been unable to deliver appropriate solutions. It drives improvements across a vast array of business and social scenarios: automation, drug research, customer relationship management, predictive maintenance, supply-chain optimization, operational effectiveness, workforce effectiveness, fraud detection, natural sciences, engineering and weather prediction. The more complex the situation, the more likely that emerging failure situations and their effective control cannot be mastered by even the smartest engineers; often, the necessary advances in transportation, energy, medicine and manufacturing would not be possible without machine learning.

Benefit Rating: Transformational

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: Alpine Data Labs; Angoss; FICO; GE Intelligent Platforms; IBM; Knime; Oracle; SAP; SAS; Skytree

Analysis By: Adib Carl Ghubril

Definition: A muscle-computer interface uses electrical signals generated from the contraction of muscles, typically in the arm and face, and detected on the surface of the skin to generate machine instruction.

Position and Adoption Speed Justification: Surface electromyography has been used in clinical biomechanics for more than 10 years, and the relation between the characteristics of the myoelectric signal and the mechanical characteristics of the movement, such as force and torque, has been studied ever since.

It is now possible to differentiate between hand gestures such as a clenched fist, a thumb-index pinch or thumb-middle finger pinch, a finger tap and lift, a wrist twist and an open palm. Indeed, the system may be configured to match the measured and conditioned myoelectric signal to a library of waveforms that correspond to preset gestures appropriate for the end application.

This system forms the basis for the operation of myoelectrical prosthetics and is now being adapted to enhance the human-machine interface (HMI) for commercial applications. These typically take the form of a forearm band equipped with surface sensors and conditioning and measuring circuits calibrated to the user and set to identify certain hand gestures. The latest applications enrich the touch experience on a surface computer and provide an alternative to visual-based, free-air gesture control.

Microsoft, along with researchers from the University of Wisconsin and the University of Toronto, built a surface computer prototype that supplements the commands generated by touch events on the screen with the signals generated from sensors strapped to a user's forearm. This expanded instruction set allows the user to change the color and thickness of a trace being rendered on screen, depending on which finger is being used and how much pressure is being applied on the touchscreen. The system also allows for gesture-based cut-and-paste commands by discerning between thumb-index pinches and thumb-middle finger pinches.

The prospects of muscle-computer interfaces are good because surface electromyograms are immune to sunlight and can detect subtle muscular changes, making them a useful complement to existing gesture control technologies. Thalmic Labs, a startup, has a product — the Myo armband — that allows users to control gaming applications using the myoelectrical signatures from preset hand gestures. The Myo is now being combined with head-mounted computers, like the Oculus Rift, to create a means for users to immerse their arms into a virtual reality application.

Myoelectric signals from facial muscles could also be used to interpret mood from facial expressions.

User Advice: Consider muscle-computer interface solutions for outdoor applications, where the object being controlled may not be within reach of the operator, such as hazardous-duty robots.

Consider also enriching the repertoire of touchscreen commands with instructions based on myoelectric signals from forearm muscles, and improving computer context awareness though surface facial electromyography — although placing sensors on the face is a disagreeable proposition that will necessitate alternatives.

Business Impact: The current touchscreen experience can be sterile; it doesn't discern a wide-enough range of tactile inputs, and it is mediocre for text entry. Supplementing it with myoelectric signals can enrich the experience and help improve the market prospects of a mobile device. The interaction between the user and the mobile device can be enhanced further by incorporating the context of mood, provided that an inconspicuous method of detecting facial muscle signals is devised.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: Microsoft; RSLSteeper; Thalmic Labs

Analysis By: Jim Davies

Definition: Emotion modulates human communication and manifests itself through facial expression, gesture, posture, tone, vocabulary, respiration and skin physiology (temperature and clamminess). Understanding what a person is communicating entails deciphering that individual's modulation scheme, which requires knowledge of social and cultural mores and cues, as well as familiarity with the individual. Emotion recognition can only lead to optimal computing if the system is able to consider all these modes.

Position and Adoption Speed Justification: Emotion detection is primarily used in call centers to identify angry or abusive customers (or agents) and alert a supervisor. In an audio recording, emotion is detected based on the analysis of what was said and/or how it was said (for example, changes in pitch, tone, frequency and volume). It's not necessary to understand what was said to determine a change in emotion. Analysis of written communications, such as email and chat, can also highlight emotion. Truth verification is being used in some call centers for fraud detection and deterrence — for example, in insurance claims. However, this is perceived as a low-priority, luxury item for many organizations, and the change management implications are too great for broad adoption. Vocal emotion can also be used to route customers to more appropriate service agents based on their current state and voice cues indicating personality profiles. Within the retail market, the emotions that organizations can strive to identify include expectation, trust, value, involvement, familiarity, danger, disgust, anger, fear, novelty, attention, desire and lust.

As organizations begin to explore and understand the customer experience across multiple channels, the online emotional state will become an important dimension to consider. Web analytics and text-mining tools will begin to piece this together, based on keystroke/mouse actions and electronic communications. Ultimately, it will be possible to determine an emotional state by body movement, as part of a gestural/3D interface, although these technologies are now associated mainly with the gaming industry. However, facial emotion detection yields more information and will be more common for communication and training, for example. Studies comparing the meaning conveyed in text, voice and visual communication have resulted in comparisons with differences as significant as 7% in actual words used, 38% in vocal tones, and 55% in facial and body language.

There are various other technologies that can be associated with emotion recognition, such as a pulse oximeter (blood flow), plethysmograph (organ volume flow), galvanic skin response (sweating), EMG (facial electromyography to monitor how the face changes), EEG (measures brain waves) and functional MRI (looks at oxygenation around the various parts of the brain using an intense electromagnetic field), eye tracking and facial recognition using infrared.

Finally, it is worth highlighting that, from a consumer perspective, machines that generate machinelike voice synthesis that cannot incorporate emotional context into responses (such as phones and cars) tend to be more inaccurate and less engaging.

User Advice: Organizations should examine whether call center applications of emotion detection could significantly improve customer service or reduce fraud. Undertake a pilot to help determine appropriateness. Contact centers that are already using speech analytics in some form should investigate the cost, impact and feasibility of adding an emotional dimension to the insights obtained from their solution provider.

Business Impact: Emotion detection can enhance the customer experience in call centers through fraud detection in intelligence, law enforcement, insurance claims and user interface improvement. Within contact centers, once supervisors have been alerted to emotional calls that are in progress, they can join the live calls and just listen in on mute, become part of the calls, or wait until the calls are over and then call the customer back. Highly emotional call recordings can be targeted for evaluation to assess how an agent dealt with the situation, and to identify any potential training needs. As the technology matures, agents will be able to alter the interaction flow based on the automatic detection of the elevated emotion and application of analytics to determine the next action, which could be anything from a refund to the escalation of the call to a supervisor.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: Autonomy; CallMiner; Genesys (Utopy); Mattersight; Nemesysco; Nexidia; Nice Systems; Verint Systems

Analysis By: Adib Carl Ghubril

Definition: Ensemble interactions refer to the idea of a user interacting seamlessly with an ensemble of different devices, maintaining synchronization across them all to support the completion of a task in a preferred and convenient manner. Ensemble interactions also apply to multichannel strategies, in which customer journeys are guided from channel to channel.

Position and Adoption Speed Justification: Human-machine interface discussions around syncing a user's various displays — via a personal cloud or a specific technology — have taken a back seat to activities around speech, "wearables" and touchless interactions through gestures and eye movements. Unfortunately, this situation hasn't changed during the past 12 months, and thus, the time for ensemble interactions to reach the Plateau of Productivity has been extended.

Although a few demonstrations of the concept have been shown, technology that bridges the physical and digital world — to allow for multitasking between paper and computer work, for example — is still immature and is the subject of academic research. However, the advent of an OS platform across computing devices of various form factors and end applications, such as iOS for smartphones, tablets and TVs, raises expectations about the concept's viability. Nonetheless, user expectations about how multiple devices should collaborate in delivering an experience are undefined and uncertain, and the number of domestic devices that could participate in interactions is limited. The growth in Internet Protocol (IP)-enabled TVs, and the increasing availability of wireless networking in computing and entertainment devices, suggests that home media will be an early area of adoption of ensemble interactions. Emerging technologies (for example, wireless video interfaces) and the use of geological location technology, which triangulates the location of devices indoors using a combination of Wi-Fi GPS and cellular signals (by companies such as Navizon), will facilitate rapid shifting among display devices.

Content providers are fundamental stakeholders in this technology and may be the most able to raise the hype on it, although they have not yet done so in any significant manner. We need to see more demonstrations from Netflix, Sony and Amazon, as well as from smart TV providers, such as LG and Samsung, on how users can, for example, continue to consume or produce content anywhere in the home by dynamically switching where the content is being displayed (that is, switching to whichever screen is available in the room or area they walk into).

User Advice: With a growing realization of the importance of the customer experience, linked to greater use of mobile and even wearable devices, user organizations should monitor this technology because, as it emerges, it could further complicate assumptions about what device an interaction is taking place on. This is because the device might change during the interaction. Organizations conducting business in situations in which two or more smart devices are in the same vicinity (such as when consumers congregate in retail stores) should pilot ensemble interactions that are implemented using communications applications.

The possibility of syncing multiple devices of many users makes ensemble interaction an enabler of collaborative work in the workplace and could also support a user's multitasking needs.

Business Impact: Vendors in areas such as mobile devices and consumer electronics should monitor and sponsor research into ensemble interactions. Vendors that manufacture a wide range of consumer electronics (for example, in the mobile and home entertainment markets) should explore this technology as a way to add value when consumers purchase more than one product from the same company. The notion of synchronizing multiple devices or multiple channels into one "ensemble" is also an enabler of big data analytics, whereby services may be tailored to user habits and preferences gleamed from a coherent assessment of user behavior.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Embryonic

Sample Vendors: Microsoft; Motorola

Analysis By: Tuong Huy Nguyen

Definition: Micro-electromechanical systems (MEMS) displays are ultra-low-power reflective displays in which pixel brightness and color are controlled by physical movement within a MEMS cell.

Position and Adoption Speed Justification: MEMS technology used for low-powered displays for mobile devices is still a relatively new technology, and deployment remains limited. There have been a number of different, competing display solutions available, primarily based on either transmissive or reflective technology, but the market has become more streamlined due to refocusing and consolidation. To reflect the low adoption of and interest in this technology, Hype Cycle positioning remains unchanged from the previous year.

Qualcomm's Mirasol solution is a reflective display technology based on interferometric modulation (IMOD) technology, with MEMS structure at its core. Qualcomm's IMOD display solution operates by using spatial multiplexing mirrors to generate color using interference and the reflection of ambient light. Qualcomm's Mirasol solution is a low-power, color, video-capable and sunlight-viewable display that has proven commercially available solutions. This solution improves on electronic ink by providing color and is able to show video/multimedia. Despite this, as a reflective technology, the color aesthetic is softer than that of standard displays because of the natural-light source. It produces a softer, less vibrant color compared with LCDs and organic light-emitting diodes (LEDs).

Furthermore, Qualcomm has demonstrated an embedded front light with LEDs that's controlled by an ambient light sensor in the e-reader solution — to improve overall usability. Qualcomm's Toq smartwatch is the latest product using the Mirasol solution. Adoption and interest of this type of device as well as wearables in general will determine the viability for MEMS displays moving forward.

We have updated the vendor list for accuracy and removed TI and Pixel Qi as neither provide a direct view MEMS display solution.

User Advice: Consumer electronics vendors should implement these solutions only if doing so improves the cost structure and utility of devices. They should also evaluate competing technologies, such as electrowetting displays from Samsung, when these are perfected. As a low-power solution, MEMS displays are applicable for handset vendors as green solutions, for developing markets (where access to electricity is limited), and as solutions to extend battery life (still a key concern for consumers). Consumer electronics vendors can also consider applications beyond the main screens on mobile phones, such as "status" screens found on consumer electronics products, and on accessories such as headsets and secondary displays on phones.

Business Impact: MEMS displays in connected portable devices are expected to be a more energy-efficient display technology than current LCD screens and to decrease costs (as they scale). This technology currently lacks scale (and requires infrastructure investment), so the benefits are offset by the added price. Moreover, it is in its early stages, with a limited number of commercially available products. With the only commercially available MEMS display solutions, Qualcomm is best-positioned for current adoption of these technology offerings. Adoption is further affected by its ability to meet demand.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: Qualcomm

Analysis By: Stephen Prentice

Definition: Behavior- or gesture-based analytics is the automated analysis of real-world human activity captured by video systems to track human movement and gestures to assess intentions and to identify specific behaviors.

Position and Adoption Speed Justification: Initial developments were targeted at the automated analysis of surveillance monitoring systems at high-risk, restricted-access locations (such as airports, government buildings, critical infrastructure, sporting events, transportation hubs and correctional facilities). The analytics capability was deployed as a part of complete systems, including multiple high-resolution cameras and recording capabilities, and deployed by selected government agencies as part of the ongoing effort against criminal and terrorist activities. These are complex systems with self-learning capabilities and extended capabilities, including facial recognition and tracking facilities.

Systems of this type are also utilized in high-value environments (such as casinos), and in the emerging physical security information management space, converging operational technologies (such as access control, alarms and sensors) with analytics information. One example is New York City's Joint Traffic Management Center, operated in conjunction with the New York State Department of Transportation.

The growing availability and usage of consumer-oriented video systems designed to support gestural control are opening up new fields of opportunity for gestural analytics. Retail is becoming a major application area, and various retailers are now actively exploring the use of video systems to measure shopper behavior and track customer movements to determine buying intentions. The rapid growth of gestural control systems in the consumer space will drive applications in the nonsecurity field at an accelerated pace, likely delivering spinoff benefits in analytics that can be used more widely. For example, Microsoft's Kinect system is experimentally tracking and recording shopper behavior to provide retail analytics, exploiting the ability to recognize and interpret body movements.

User Advice: The high cost of integrated surveillance systems limits their deployment by government agencies to high-risk hot spots; but ultimately, the reductions in staffing levels, combined with growing capabilities to detect subtle behavioral movements, make this a powerful and attractive technology, although there is still much development to do. Security and anti-terrorism demands will overpower the concerns of civil libertarians regarding the "surveillance society" in public spaces. In commercial environments, only the most high-risk areas can justify coverage of this nature.

However, interest in and availability of behavior- and gesture-monitoring systems targeted at nonsurveillance markets are increasing. These widely varied systems provide a rich stream of data to combine with other sources to better understand human behavior and predict future intentions. Market research, commercial real estate (sporting venues, shopping malls and so on) and retail are all areas where a thorough investigation of the emergent capabilities of this technology is easily justified.

Business Impact: The ability to automatically monitor human behavior in public spaces and detect the warning signs of undesirable activities is very compelling in the security and risk management areas, border and customs agents, and government sector. The potential for using the same capabilities to understand activity and predict future intentions also has huge implications in most retail and commercial environments.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: IBM; Mattersight; Shopperception; VideoMining

Analysis By: Gene Phifer

Definition: A user experience platform (UXP) is a rationalized, integrated set of technologies used to provide interaction between a user and a set of applications, processes, content, services or other users. While similar, a UXP is different from a portal product in that it contains a broad collection of supporting services that typically live outside the scope of a traditional portal product. These services include portal, app/API framework, content management, search, analytics, collaboration, social, mobile and UX framework.

Position and Adoption Speed Justification: The concept of the UXP emerged in 2009 and UXPs are still in their infancy. A UXP is, first and foremost, a platform for enterprises to use. They may create the platform themselves or purchase it as a suite of products or as a single product. There have been early examples of UXP offerings (such as Microsoft SharePoint, Oracle WebCenter, IBM Customer Experience Suite and Employee Experience Suite, Adobe Marketing Cloud and Creative Cloud, and Oxcyon); however, the UXP is in the early phases of its life cycle. The UXP market solidified in early 2014, the UXP market will grow as vendors promote their UXP offerings as the best path to customer, employee and partner engagement.

The optional components of the UXP will evolve over time. For example, additional components to facilitate digital marketing will likely be added to address the digital marketing use case. In fact, strong customer versus employee use cases may drive different technology sets, and may cause market segmentation of the UXP over time.

User Advice: Conduct an inventory of the various tools used for presentation management and presentation-layer composition across all supported devices and channels. Determine synergies where common vendors are identified. Demand plans from your vendors for their products, and determine where integration will occur. Explore UXP options for many of these tools, then build a road map and plan to adopt a UXP during the next few years.

Determine whether your needs for websites and portal sites would best be met by suite-oriented portal products/UXPs or by lean portals.

If you already have most of the components of a UXP and are happy with them, fill in any gaps and continue with a "roll your own" approach. If you are lacking major components, consider a UXP product as the source for those missing components. If you don't have much of a platform, or don't like most of the components you are using, consider a full UXP product from a UXP vendor.

Business Impact: Tools and methodologies for delivering the user experience can be expensive. However, the biggest downside is the impact on the users who have to deal with inconsistent user experiences and a different look/feel/behavior across different sites and devices. A UXP provides significant efficiencies in developing and maintaining the user experience, and provides a consistent user experience across sites, channels and devices. UXPs are likely to save organizations money, but the main benefit is the ability to more effectively and rapidly engage users and customers.

The UXP addresses the enterprise need for a consistent, integrated, versatile and optimized approach to user interactions across a wide range of scenarios (business-to-employee [B2E], business-to-consumer [B2C], B2B, government-to-constituent [G2C] and digital marketing) and devices (PCs, smartphones, tablets, consumer electronics, appliances, cars and a variety of emerging devices, such as wearables). The preintegrated nature of UXP products means faster time to market and lower deployment costs. These cost savings are offset, to a degree, by relatively high licensing costs. Because of the large investment in money and resources required of an on-premises UXP, small enterprises and some midsize enterprises should explore cloud UXPs or lean portals as alternatives.

Benefit Rating: High

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: Adobe; eXo; IBM; Liferay; Microsoft; OpenText; Oracle; Oxcyon; salesforce.com; SAP

Analysis By: Thilo Koslowski

Definition: Mood recognition technologies sense the emotional state of a user (via biometric sensors, including in fabrics like seats or clothes, cameras, and interactions) and respond by performing specific, predefined actions, such as changing the lighting in a vehicle to more-subtle colors to address a user's high-stress level or playing dynamic music to address driver fatigue. The technology can also be used for other applications and segments, including driver education to reduce potential accidents and improve the driving experience.

Position and Adoption Speed Justification: Mood recognition technology has not yet found its way into many production vehicles. Vehicle manufacturers, suppliers and technology companies continue to explore the potential of this technology and focus efforts on the most effective ways to market such features in the future. Growing interest in addressing driver distraction is creating more awareness for mood sensing, but the technology is still far from broad commercialization. The first broad applications of mood-recognition-based automotive applications are likely going to be infotainment-centric solutions, such as playing specific music based on traffic flow, driver attention and other contextual factors.

User Advice: Continue to explore the potential of mood-recognizing technologies in automobiles. With the increased demand for a driver's cognitive attention, and because of increased traffic situations and the emergence of new data services, mood recognition technologies can provide relief as well as minimize driver stress and fatigue. Partner with automotive suppliers, consumer electronics companies, universities and behavioral experts to further explore this field. Observe and adapt from other industries where applicable (for example, retailers and advertising forms are experimenting with the technology).

Business Impact: Mood-sensing technologies provide differentiation, especially for specific customer segments (for example, business drivers, long-distance commuters and new drivers). Furthermore, the technologies will improve driver attention in traffic and during interaction with data services in a vehicle.

Benefit Rating: Moderate

Market Penetration: 1% to 5% of target audience

Maturity: Embryonic

Sample Vendors: Affective Media; IBM

Analysis By: Adib Carl Ghubril

Definition: Speech-to-speech translation involves translating one spoken language into another. It combines speech recognition, machine translation and text-to-speech technology.

Position and Adoption Speed Justification: Speech-to-speech translation entails three steps: converting speech into text; translating text; and finally, converting text to speech. In effect, anything that may be converted to text may be translated. Microsoft's Bing platform, running on Windows 8.1, and Google Translate offer speech translation; furthermore, their optical character recognition middleware allows the user to touch or select an on-screen character, from a graphic or photo, and listen to the description or translation of that resulting text in any of the supported languages.

While there has been little adoption of the technology by enterprises to date, due to accuracy limitations and response times, the availability of low-cost mobile consumer products may drive interest and progress for higher-end applications. We continue to anticipate rising hype and capabilities during the next two years, and a growing breadth of applicability during the next five years. In August 2013, Facebook purchased Mobile Technologies, makers of the speech-to-speech application "Jibbigo," and this acquisition is a reflection of Facebook's ambition to enable online interaction.

Vendors can build on their speech recognition know-how (such as what Apple has done with Siri), to create a translation system that can be used to support dialogue. Meanwhile, platform-specific applications from independent developers — like SayHi Translate — continue to expand the breadth of users' options. Also, experiments are being conducted with a multimodal approach, in which information from gestures and facial expression is being used to execute translation in context with dialogue. IBM is tackling out-of-vocabulary words by devising a machine that interacts with the user to ascertain where the linguistic mistake was made.

User Advice: Do not view automated translation as a replacement for human translation but, rather, see it as a way to deliver approximate translations for limited dialogues in which no human translation capability is available. Evaluate whether low-cost consumer products can help during business travel or first-responder situations. Leading-edge organizations can work with vendors and labs to develop custom systems for constrained tasks.

Business Impact: Consumer mobile applications are the first to attract significant interest. Potential enterprise applications include on-site interactions for fieldworkers, as well as government security and emergency and social service interactions with the public. In the U.N. General Assembly, more than 20 languages are spoken, and the yearly meeting transcription fees are significant — finding ways to automate that would provide welcomed cost relief.

Speech-to-speech translators can help improve the social interaction between foreign soldiers and local inhabitants in the urban settings of modern-day theaters of war. In the longer term, multinational call centers and internal communications in multinational corporations will benefit, particularly for routine interactions. However, internal collaborative applications may be limited because strong team relationships will unlikely be forged, if the only way to communicate is through automated translation.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Cellictica; Facebook; Google; IBM; Microsoft; Philips; Science Applications International Corp.; SpeechTrans

Analysis By: Carl Claunch

Definition: This computing system architecture has processors that use more than one instruction set, all of which share a single memory. This requires programs to be written differently for each of the dissimilar instruction sets. The goal is to offer substantially better performance or cost by devoting the appropriate parts of the application to machine designs optimized for specific types of computing.

Position and Adoption Speed Justification: This concept has been established in specialized, embedded computing systems. It is used in areas such as telecommunications and graphics processing, where separate machine architectures provide substantial benefits to relevant parts of the overall task. This yields higher overall performance than if a homogeneous system were designed.

The operating system must manage the collection of machine elements as a single system while dispatching the independent queues of work written to each instruction set. Heterogeneous architectures are available in partial implementations that use graphics cards (general-purpose GPUs [GPGPUs] or GPU computing), field-programmable gate arrays (FPGAs) and other components with direct-access drivers, in lieu of full support from the operating system. The processors can all be targeted to run parts of an application by creating appropriate code. If the processors are restricted to specific functions, such as XML processing, then they are covered in the profile of offload engines instead.

Progress in this area in 2013 came from many providers, including IBM (common API [CAPI] heterogeneous memory sharing in Power8), Nvidia (GPUDirect in Kepler architecture), and AMD (Heterogeneous System Architecture [HSA] heterogeneous memory architecture). The awareness of potential use of this technology in general computing applications has increased the most this year, catching up with high-performance computing market interest levels.

User Advice: In the near term, users with extreme requirements — those that are not satisfied with the expected future capabilities of homogeneous systems and those that can justify the risk and investment required to experiment with heterogeneous systems — should consider projects based on this technology. Others should keep an eye on progress in this area, looking more deeply at the concept once it has moved to a more mature state and is further along the Hype Cycle. Systems with mixed architectures are likely to cost more than monolithic x86 servers. The benefits from increased performance or lower total numbers of servers should be enough to offset the added per-system cost, as well as the added complexity of dealing with heterogeneous machines and software.

Business Impact: Heterogeneous architectures may offer previously unattainable levels of performance for HPC-like workloads. However, they will eventually become a mainstream design for servers in the general commercial market.

Benefit Rating: High

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: AMD; Cray; HP; IBM; Intel; Nvidia; SGI

Analysis By: Michele Reitz

Definition: Sensor fusion refers to algorithms that aggregate and transform many disparate sensor inputs to provide useful outputs to a device or system. A sensor fusion solution usually combines a hardware sensor "hub" function and a software sensor fusion stack. However, in some cases, hardware is purpose-built to perform full sensor fusion.

Position and Adoption Speed Justification: Aggregation of sensor inputs into a sensor hub for automotive and industrial systems has been prevalent for over 10 years, and this approach has been used in smart TVs and video game consoles for the past five years.

With the rapid changes in the consumer market, specifically in mobile devices, sensor fusion has evolved and is now at the forefront of new integrated functionality in smartphones, tablets and new markets contributing to the Internet of Things. In the past two years, we have seen typical high-end smartphones each include over 10 sensors: inertial sensors (accelerometer, magnetometer, gyroscope, barometer), three microphones, two cameras, one light sensor, one proximity sensor, two touch sensors, other biometric sensors, and many types of radio (cellular, GPS, Wi-Fi, Bluetooth, Near Field Communication). Sensor fusion gathers inputs from these sensors and formulates outputs that indicate the phone user's activity, position, movement and context. This information is then used to feed apps.

In addition, a discrete low-power sensor hub solution can offload the main processor, thereby reducing power consumption and enabling "always on" monitoring of sensor inputs. One of the most hyped sensor hub solutions was announced by Apple for its iPhone 5s, in which NXP's M7 coprocessor chip offloads the A7 application processor to perform sensor monitoring and fusion activities. Google's Nexus 7 device uses a sensor fusion solution from InvenSense to make motion calculations. Samsung's Galaxy S4 device uses an Atmel sensor hub coupled with sensor fusion software from Hillcrest Labs. Samsung's Galaxy S5 devices also use a sensor fusion solution from InvenSense.

Sensor fusion solutions developed for mobile devices have been the most notable. They have also formed the foundation of a variety of applications, including new types of wearable computing devices, fitness and health-monitoring devices, video game devices, indoor navigation devices, drones, connected automobiles, and other devices in the Internet of Things. As consumer device companies continue to add more and different types of sensors to differentiate their products, sensor fusion is foundational to efforts to interpret all the different sensor inputs by doing the "smart" work required to transform them into useful information.

A robust sensor hub solution will need "always on" functionality, interfaces to the host processor (if there is one), sensor calibration and sensor drivers as part of the sensor software stack. Sensor fusion software stacks are available from companies such as Hillcrest Labs, Sensor Platforms and Kionix. They provide libraries that can be ported to a variety of different hardware sensor hub platforms. Sensor hub hardware in the form of application processors, microcontrollers, programmable logic and other purpose-built devices have been introduced by companies such as Atmel, Freescale, InvenSense, Lattice Semiconductor, QuickLogic and STMicroelectronics.

In view of the increasing use of sensor fusion/hub solutions in smartphones, as well as the large number of software and hardware companies with related products, we consider this technology to be adolescent, and we place it very near to the Peak of Inflated Expectations.

User Advice: Users should evaluate the various hardware and software approaches to determine the right balance of power savings and cost for each application, and the complexity of the operations needed for their devices.

Business Impact: Sensor fusion technology can achieve significant power savings, enabling longer device operation times and the use of smaller batteries, which reduces the overall bill-of-materials cost.

In addition, clever manipulation of input sensors to provide information about device usage and user context creates opportunities for new consumer apps and services.

Benefit Rating: Moderate

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: Apple; Atmel; Hillcrest Labs; InvenSense; Kionix; Lattice Semiconductor; Microchip Technology; QuickLogic; Sensor Platforms; STMicroelectronics

Analysis By: Whit Andrews; Hanns Koehler-Kruener; Tuong Huy Nguyen

Definition: Natural language question answering (NLQA) technology is composed of applications that provide users with a means of asking a question in plain language. A computer or service can answer it meaningfully while maintaining the flow of interaction.

Position and Adoption Speed Justification: The challenges in effective interpretation of idiomatic interrogative speech, matching it to knowledge bases of potentially infinite scope, and the selection of a limited number of answers (even just one) remain profoundly difficult. Simple answers such as the one answer available for a trivia question are far easier than the multivariate, nuanced answers inherent in real human communication (for example, "Cold or flu? Why not cold AND flu!").

IBM captured the attention of the world in February 2011 when Watson (a Smart Advisor) won a quiz show, and now the technology is maturing into a variety of sophisticated products. It joins a long line of immediately fascinating and broadly constrained custom-made knowledge-calculation devices. This has been followed by the mainstream introduction of simple conversational assistants from Apple, Google and, most recently, Microsoft. Apple's Siri launched in 2011 as a new way for users to interact with informational systems. It incorporates speech-to-text technology with natural-language processing query analysis to wow users (at least some of the time). In 2013, Google featured such technologies in the keynote for its closely and avidly watched Google I/O. Such attention defines a peak of hype. Facebook's Graph Search project also allows for semantically rich queries, albeit with less ambiguity. In April 2014, Microsoft introduced Cortana: a virtual personal assistant for the Windows Phone operating system. As precursor products, these conversational assistants should evolve into virtual personal assistants between 2015 and 2017.

Solutions ultimately must discover means of communication with humans that are intuitive, effective, swift and dialogic. They benefit significantly from context, either detected (as in geographical location) or explicit (as with products that have a specific goal, such as health diagnostics). The ability to conduct even a brief conversation, with context, antecedent development and retention, and relevancy to individual users is in its infancy. However, nonconversational, information-centered answers are indeed already possible with the right combination of hardware and software. Also, as in all technology categories, the availability of such resources can only become cheaper and easier. More than five years will pass before such capabilities are commonplace in industry, government or any other organizational environment — but they will ultimately be available to leaders in such categories.

User Advice: The computing power required to accomplish a genuinely effective trivia competitor is expensive, but will become less so with time. Any projects founded on such facility must be experimental, but in the foreseeable future will include diagnostic applications of many kinds as well as commercial advice and merchandising, and strategic or tactical decision support. "Augmentation" of human activity and decision making is the key thought. No decision support application comes, fully formed, from nothing — it will be expert humans who build it, design the parameters and develop the interface. Humans will, similarly, evaluate its advice and decide how to proceed. A good idea is to begin with experimental technologies, such as chatbots, and to work toward more sophisticated technologies as they become commercially accessible.

NLQA is positioned to be a strong enabler for other technologies such as virtual personal assistants, cognitive computing and speech recognition. These technologies can serve as two-way steppingstones toward building an effective NLQA system.

Business Impact: Ultimately, the ability for line workers or unschooled consumers to achieve effective responses from machines without using expertise in framing queries will generate new kinds of information exploitation by diminishing information friction yet more. Given a limited set of answers and an effective means of capturing plain language requests, it is easy to see computers more effectively providing guidance in various environments. Business cases such as diagnostic support in healthcare — whether for expert or nonexpert users — and also consumer services (such as those Siri provides) are some use cases.

Benefit Rating: High

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: Apple; Cognitive Code; EasyAsk; Expect Labs; HP; IBM Watson; Microsoft; Nuance; Sherpa Software; Vlingo; Wolfram Alpha

Analysis By: Angela McIntyre

Definition: Wearable user interfaces describe the interaction between humans and computing through electronics designed to be worn on the body. They may sense the human body or the environment around the wearer and transmit the information to a smartphone or to the cloud. Ideally, wearable user interfaces are unobtrusive, always on, wirelessly connected and provide timely information in context. Examples of wearable electronics are smart watches, smart glasses, smart clothing, fitness monitor wristbands, sensors on the skin and audio headsets.

Position and Adoption Speed Justification: This past year saw the hype on wearable user interfaces reach the Peak of Inflated Expectations and become tempered with realism about the value consumers perceive in new wearable devices. Devices are not viewed as stylish by consumers; the data collected from wearable sensors yields insights that are marginally useful to wearers. Apps and algorithms that can interpret noisy data from wearable sensors are needed to increase the usefulness of recommendations. Use cases in which wearable user interfaces are more convenient than smartphones are limited. Smartphones are gaining sensors and apps that give them health and fitness tracking capabilities similar to wearables. Nonetheless, apps on smartphones are enabling new capabilities and insights from wearables.

Yet interest in wearable interfaces remains high, and Google is fostering an ecosystem that is expected to gain traction. Android Wear will enable a consistent user interface across different types of wearables. For example, Android-based wearables will have a similar user experience in the layout of glanceable displays, gestures for navigating among apps, and using voice commands for control and to access information. Google and affiliated service providers have potential access to personal information gathered by other wearable devices using services on an Android Wear platform.

Data from wearable devices can be combined with data from other devices in the Internet of Things and from other sources on the Internet, adding to big data. Apps, services and virtual personal assistants (VPAs) will provide increasingly useful insights to wearers as part of cognizant computing by using personal data collected from wearable electronics. The consumer trends for adoption are being driven by quantified self, convenience and the desire for immediate alerts, especially regarding social networks. Tracking data for medical reasons will be a longer-term driver for wearables.

Over the next 10 years, wearable user interfaces will enable services to become more personalized to the preferences and needs of the user through contextual information and bio-data gathered through wearable electronics. Similarly, wearable devices will serve as controllers for other devices in the Internet of Things. For example, consumers with Nest thermostats can control them remotely through Google Glass. Similarly, the Pebble smartwatch can take a photo with the GoPro camera and start a car engine remotely. These are early examples of how wearable user interfaces will become increasingly integrated into daily life.

User Advice: Invest now in deployments or pilots for wearable user interfaces in the enterprise. Start with wearables for mobile workers who cannot conveniently or safely put aside what they have in their hands to use a phone or tablet, such as employees using tools or equipment, or who need to keep their heads up or to hold on for safety.

Engage with software developers now on augmented reality use cases specific to your business needs. Augmented reality solutions are in development for head-mounted displays (HMDs). However, robust software solutions using augmented reality beyond checklists will take an additional two to five years of development. The battery life of present HMDs lasts only a couple of hours for uses such as streaming video. Until at least full-day battery life is available, workers will find wearables inconvenient or impractical to use.

Where time-motion efficiency is essential to productivity, such as in call centers and logistics organizations, employers are investigating wearables, such as gaze tracking through audio headsets and location tracking through badges. Explore solutions that lead to recommendations to increase worker productivity or to monitor employees in physically demanding work environments.

Encourage the workforce to be healthier by implementing wellness programs that include wearable fitness trackers and also work with providers on advances in algorithms for fitness trackers. Fitness trackers in wristbands or other forms are motivating to people who want to be less sedentary. The general health of the consumer or employee can be measured with wearables, including body temperature, heart rate, heart rate variability (stress) and potentially blood glucose.

Explore longer-range opportunities for always-on information access through smart glasses or smart watches through voice input and video, but evaluate risks before heavily investing. Create policies around personal privacy and the restrictions around taking pictures in the workplace. Data security risk will also increase with the rise in content sharing among devices that are interacting across personal networks.

Business Impact: Early industries to adopt wearable electronics are aerospace and police, followed by sports, field service, manufacturing, logistics, transportation, oil and gas, retail, and healthcare. The healthcare market stands to benefit from wearable user interfaces that enable mobile health monitoring, especially for heart conditions. Wearable cameras are ready for deployment now for use cases such as police/security and inspections. Field service and manufacturing are using streaming video to an expert who sees what the wearer sees, which is useful for training or expert assistance. Sports is using wearables on players for "in the game" perspective tracking the performance of athletes. Augmented reality solutions on HMDs have the promise to increase productivity by providing part labels, checklists, maps and instructions superimposed on real-world views.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Aliph; Eleksen; Epson; Eurotech; Fitbit; FitLinxx; General Dynamics Itronix; Google; Kopin; LXE; Motorola; Oculus VR; Pebble; Plantronics; Recon Instruments; Samsung; Sony; Vuzix

Analysis By: Anshul Gupta

Definition: Smart fabric refers to a range of technologies that transform material used in clothing, upholstery and other textiles into devices that can be deployed as sensors, switches, connectors, batteries or displays. The components and electronics may be embedded in the fabric or, in some cases, on the fibers themselves.

Position and Adoption Speed Justification: Smart fabrics are being used for a wide range of applications:

• Controllers for electronics
• Human physical data monitoring
• Alarm systems
• Heating wraps
• Energy harvesting
• Sport monitoring (by professional and nonprofessional athletes)
• Electromagnetic shielding and illumination

Smart fabrics with embedded solar cells or piezoelectric panels are being used to collect, conduct and store energy, which can be used to power electronic devices such as mobile phones and provide off-grid power solutions. Cloth with embedded sensors can measure and store data on temperature, heart rate, blood oxygenation, blood glucose, position, motion, moisture, chemicals, respiration, light, pressure and more. Smart cloths can be made to emit light. For example, luminescent fibers can be woven into cloth used for curtains, clothing and wallpaper, enabling them to glow in a range of colors that can be seen in a darkened room. In the automotive industry, smart fabrics can transform the material used in car seat covers, body linings, panel coverings and other fabric goods into devices that can be deployed as electronic sensors or switches.

Data can be transmitted wirelessly (Bluetooth) by powering antennas and sensors through a thin-film battery.

However, adoption of smart fabric will be slow until 2017, due to technology's immaturity and initial adoption in niche areas such as fitness wear, art lobbyists, heated winter wear, garments for charging and controlling MP3 players, solar tents and medical wraps. Applications of smart fabric in the many areas discussed above are in early stages of development, and significant improvements are needed before mass commercial adoption. Some of the key areas of improvement are manufacturing economically, delivering high quality, cleaning fabric without affecting their electronic components, efficiency, privacy concerns and user awareness.

Smart fabric will provide immense opportunities for enterprises with innovations in the technology. Service providers in automotive, healthcare, military, emergency services, fitness clothing, mining, engineering and manufacturing will be most impacted by innovation in smart fabric technology.

User Advice: Providers of smart fabric products should:

• Focus on designing easy-to-use, simple products with concentration on maximizing benefits over costs.
• Consider regulations surrounding medical products — especially, for example, the Health Insurance Portability and Accountability Act from the U.S. government.
• Consider building waste-preventive measures at an early phase of the smart fabric development process, to minimize cost and reduce the risk of slow adoption.

Business Impact: Fabric interfaces and displays are driven by sporting, fashion, aesthetic, power or mobility needs. They will also play a key role in the development of smart cities, most likely emerging first in the areas of emergency services and healthcare where highly mobile, responsive technology will have increased value. Solar panels woven into draperies, patio awnings and upholstery will generate enough electricity to power automatic blinds and handheld consumer electronics.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: Eeonyx; Fibretronic; FTL Global Solutions; Peratech; Philips Research; Textronics

Analysis By: Brian Blau

Definition: Head-mounted displays (HMDs) are small displays or projection technology integrated into eyeglasses or otherwise worn on the head. Heads-up displays (HUDs) are a type of HMD that does not block the user's vision, but superimposes an image on the user's view of the real world. An emerging form of HUD is one where the display is either integrated into or paired with contact lenses. In all cases, the user perceives the virtual image at an ideal viewing distance, although no screen is present.

Position and Adoption Speed Justification: HMDs are more popular in 2014 than at any point in the past, mainly driven by Google Glass and Facebook's acquisition of Oculus Rift in April 2014. Prior to 2014, HMDs were mainly deployed in military applications (such as driver, flight and combat training) where the technology is well-developed, but has yet to find a vehicle for broad adoption in consumer markets. Several HMDs have targeted the gaming and video markets but with little success. Availability of stylish, consumer-grade video eyeglasses may eventually drive adoption; however, most attempts to launch products in this market have failed to take hold.

There are several notable new product initiatives that could push HMDs into the spotlight. In 1H12, Google Glass was announced, a wearable computing device that includes a small HUD viewing element and integrated camera mounted on simple frames. In April 2014, Google launched Android Wear, an OS and developer API meant to support Google Glass-like wearable devices. In late 2012, Oculus Rift landed a $2.4 million Kickstarter campaign, one of the largest at the time, to develop a fully immersive HMD that is targeted at the video game industry. In an unexpected move, Facebook acquired Oculus in April 2014 for $2 billion. In March 2014, Sony announced Project Morpheus, which is their HMD and virtual reality (VR) effort that is targeted toward console gaming. Sony, Oculus and a host of new HMD entrants, such as Carl Zeiss, Vuzix, Sensics, Meta and Cybermind, will soon compete over consumer- and business-grade HMD devices.

Despite early enthusiasm for Oculus Rift and Google Glass, renewed interest in HMD technology in the broader consumer market appears to be gaining momentum. Developers have responded by ordering more than 70,000 Oculus Rift prerelease kits, which means there will be many developers creating VR solutions over the coming years. However, given the significant user experience barriers and lack of associated software services, we expect continuing disillusionment with HMD technology for the next few years. This is certain to change course when HMD get into users' hands and VR creates realistic and compelling immersive experiences. Otherwise, users could suffer motion sickness or have other reactions that would limit their ability to use VR. On the positive side, growing popularity of augmented reality on mobile devices may provide some crossover and momentum for similar functionality on a head-mounted form factor. In the future, we will see continued improvements to HMDs that will focus on higher-resolution display, expanded field of view, better battery life, comfort and lower cost — all of which are constraining current adoption.

User Advice: Consider HMDs for wearable or augmented reality applications, when the user's hands are occupied with a task or when the user is moving while accessing information — for example, to review work instructions, schematics or customer data. Content and service providers should monitor Google Glass and Oculus Rift and be prepared to plan for integration, but they must consider factors like user experience, being first to market, and finding business models that support product and market growth.

Business Impact: Virtual reality will have its biggest impact when consumer-grade device sales start, and it will first impact the entertainment market through games and immersive experiences. Aside from simple video viewing, impact areas will be in vertical markets, when combining key information with real-world views improves safety and productivity (such as in medical or engineering environments). Business uses will be in mobile communication, computing devices, simulation and training, business simulation, remote worker and telepresence business processes, and, in general, will support context-aware computing interfaces that benefit from the immersive experience that virtual reality provides.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: Google; Innovega; Oculus VR; Rockwell Collins; Sensics; Vuzix

Analysis By: Adib Carl Ghubril

Definition: Facial recognition is the process of identifying humans by detecting their face and matching a scanned image of their face to an entry in a database of known samples using feature extraction algorithms.

Position and Adoption Speed Justification: Although the main thrust in facial recognition started in the early 1990s, the technology, until recently, has not enjoyed much adoption beyond law enforcement application. This situation was not due to a lack of algorithm development: Indeed, work-around facial recognition algorithms have made considerable strides, maturing from a vector-based to a neural net-based approach. Rather, it is due to the unresolved challenges that invariably defeat any such algorithm.

At the top of the list of challenges is illumination. Ambient light conditions dramatically affect how edges, features and texture are perceived. Pose, or the orientation of the face, relative to the capturing device also affects the results generated by facial recognition algorithms because these algorithms are all designed based on a full-frontal view of the subject's face. Occlusions, such as beards, glasses, hats or physical structures coming into the field of view as the subject walks around, cause partial capture of a subject's face, which weakens the recognition system. Finally, facial expressions, such as a grimace, smile or frown, also introduce variability in the analysis results. In some cases, multiple pictures must be taken of the same face to collect enough reliable data points to generate a match with a reasonably degree of certainty.

Nevertheless, adoption is accelerating, and the technology is finding applications in markets such as access management, biometrics and home surveillance. However, facial recognition is capturing the most attention in digital image editing, wearable electronics and gaming. Facebook is running facial recognition algorithms, via its "DeepFace," to help identify people in pictures. DeepFace constructs a 3D model of people's faces from the 2D data available in the picture and then applies deep learning techniques to match the resulting pattern to known good entries in its database, and the machine is reported to be accurate: Its ability to perform that photo-matching activity is equivalent to the ability of a human to perform that activity.

However, the use of facial recognition technology by government security agencies is stoking public anxiety about living in a world where privacy does not exist and in which a policy to mitigate the risks of a privacy breach is not forthcoming. Resolving the challenges and sorting out an agreeable arrangement for utilizing facial recognition in public — one that doesn't infringe on personal rights — are paramount to the progression of the technology.

User Advice: Use facial recognition as a more convenient alternative to passwords for employees requiring access to computers or entry to controlled areas. Governments and security organizations can use facial recognition to identify known criminals or suspects from video recordings.

Business Impact: Improving a system's rigor and raising its level of automation enrich the system's interaction with humans, which, in turn, raises productivity and improves adoption as users seek to repeat those particular experiences.

Facial recognition is a powerful enabler for both automating access systems in vehicles or buildings and providing more rigorous identification in areas such as passports or voter registration. It is also a powerful differentiator when developing collaborative games or sharing photos on social media sites.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Cognitec Systems; Eurotech; Facebook; Herta Security; Megvii

Analysis By: Tuong Huy Nguyen; CK Lu

Definition: Augmented reality (AR) is the real-time use of information in the form of text, graphics, audio and other virtual enhancements integrated with real-world objects and presented using a heads-up display (HUD)-type display or projected graphics overlays. It is this "real world" element that differentiates AR from virtual reality. AR aims to enhance users' interaction with the environment, rather than separating them from it.

Position and Adoption Speed Justification: Although verticals such as automotive and military have been using AR for many years, this technology entered the mainstream driven by the interest and proliferation of mobile devices and geolocation services. Recent focus has shifted back to vision-based identification AR. This technology supplements location-dependent AR and provides additional use-case scenarios.

A growing number of brands, retailers, manufacturers and companies in various verticals have shown interest in, or are using, AR to enhance internal and/or external business processes. Hype around AR has stabilized. This has allowed more companies to look beyond the initial hype to explore AR's potential to provide business innovation, enhance business processes and provide high value to external clients. The biggest challenge for external-facing AR is gimmicky implementations — solutions that provide the consumer no value. This will potentially limit consumer interest and adoption in the technology. Internal-facing implementations have better potential for adoption to bring business value because they won't be hindered by consumer preferences. Advancement of heads-up display will further encourage use of AR as an enterprise tool.

Beyond audience-based challenges, a number of factors will continue to hinder AR adoption:

• Rigorous device requirements restrict the information that can be conveyed to the end user. Cloud computing initiatives will alleviate some of this burden.
• Data costs for always-on connectivity.
• Privacy concerns for both location and visual identification-based AR.
• Standardization for browsers' data structure.

User Advice:

• Communications service providers: Examine whether AR would enhance the user experience of your existing services. Compile a list of AR developers with which you could partner, rather than building your own AR from the ground up. Provide end-to-end professional services for specific vertical markets, including schools, healthcare institutions and real estate agencies, in which AR could offer significant value.
• Mobile device manufacturers: Recognize that AR provides an innovative interface for your mobile devices. Open discussions with developers about the possibility of preinstalling application clients on your devices and document how developers can access device features.
• AR developers: Take a close look at whether your business model is sustainable, and consider working with CSPs or device manufacturers to expand your user base; perhaps by offering white-label versions of your products. Integrate AR with existing tools, such as browsers or maps, to provide an uninterrupted user experience.
• Providers of search engines and other Web services: Get into AR as an extension of your search business. AR is a natural way to display search results in many contexts.
• Mapping vendors: Add AR to your 3D map visualizations.
• Early adopters: Examine how AR can bring value and ROI to your organization and your customers by offering branded information overlays. For workers who are mobile (including factory, warehousing, maintenance, emergency response, queue-busting or medical staff), identify how AR could deliver context-specific information at the point of need or decision.
• Brands, marketers and advertisers: Use AR to bridge your physical and digital marketing assets and drive increased engagement with your user base.

Business Impact: AR is used to bridge the digital and physical world. This has an impact on both internal- and external-facing solutions. For example, internally, AR can provide value by enhancing training, maintenance and collaboration efforts. Externally, it offers brands, retailers, marketers and the ability to seamlessly combine physical campaigns with their digital assets.

CSPs and their brand partners can leverage AR's ability to enhance the user experience within their location-based service (LBS) offerings. This can provide revenue via set charges, recurring subscription fees or advertising. Handset vendors can incorporate AR to enhance UIs, and use it as a competitive differentiator in their device portfolio. The growing popularity of AR opens up a market opportunity for application developers, Web services providers and mapping vendors to provide value and content to partners in the value chain, as well as an opportunity for CSPs, handset vendors, brands and advertisers.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Adolescent

Sample Vendors: Catchoom; Daqri; GeoVector; Google; HP; Layar; Metaio; Mobilizy; Nokia; Qualcomm; Tonchidot; Total Immersion; Zugara

Analysis By: Adib Carl Ghubril

Definition: An embedded graphical user interface (GUI) provides icons, and other visual cues, to control machines with less available processing and memory resources for a more rich-interaction environment.

Position and Adoption Speed Justification: Embedded GUIs have run typically on Windows CE and embedded Linux — both of which entail a significant amount of memory external to a midtier processor. The introduction of Java Platform, Standard Edition Embedded (Java SE Embedded) started the migration of C++ widgets to JavaScript, and so more powerful functionality was able to run on more-limited resources, such as those found on board microcontrollers. However, JavaScripts can consume much of the limited onboard resources and restrict the functionality of an embedded Web browser GUI.

HTML5 is much more code-efficient, so the implementation offers developers the flexibility to customize further down the application stack. The announcements of various development toolmakers, such as mbed, Bsquare and QNX, bode well for the progression of HTML5, with the caveat that security and standardization across devices remain elusive and pose the gravest risk to reaching the Plateau of Productivity.

Indeed, the first version of HTML5 is to be completed within the next two years and, at the onset, may not support pure HTML5 applications but, rather, require wrapped HTML hybrids. QML is also a competitive alternative to JavaScripts, and vendors such as Renesas Electronics and STMicroelectronics offer their users embedded GUI development platforms. However, there has been no compelling progress on keeping the GUI stack light while maintaining a user experience that is similar to the experience a user has with commercial platforms. The integration of the GUI development platform with the general development platform (similar to what Micrium offers) is key to crossing the Trough of Disillusionment.

User Advice: Perform a cost-benefit analysis on the implementation of widgets and WebSockets using the main programming and scripting language.

Business Impact: The Internet of Things is fueling the proliferation of embedded processors and microcontrollers. The Internet of Things running HTML5 on microcontrollers not only will maintain a good power-performance ratio, but also will offer a compelling user experience via GUI Web browsers.

Benefit Rating: Moderate

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: Bsquare; mbed; Micrium; QNX; Renesas Electronics; STMicroelectronics

Analysis By: Amy Teng

Definition: A pico projector is a very small projector module that can be integrated into a small or mobile device. It can be embedded in a smartphone, the near-eye see-through display in smartglass, the overhead display of a car, or the stereolithographic apparatus (SLA) of a 3D printer. Pico projectors can be implemented using several technologies, including liquid-crystal-on-silicon (LCoS) imager, digital micromirror array chip, or a line scanning system using a light-emitting diode (LED), laser diode or other light sources.

Position and Adoption Speed Justification: "Smartphone-embedded projector" was the original target market of pico projectors. The first-generation modules small enough to embed in a mobile phone emerged when Samsung introduced its Galaxy Beam projector phone at Mobile World Congress (MWC) 2009 (Asus had previously demonstrated a laptop using a pico projector at Computex 2008). It then introduced a new iteration of the Galaxy Beam with better specification, longer battery life and slimmer design at MWC 2012. During the past year, Sony has also established its PJ family of digital video cameras that are equipped with pico projectors.

2013 saw pico projector modules lose the thickness competition with smartphones. Being embedded in smartphones becomes problematic because smartphones thin at a faster pace, and the features of pico projectors are not especially attractive because tablets are now widely available with larger displays that can fulfill a small group's need to share images and video.

Consequently, pico projectors have diversified into new emerging applications, including Google Glass and 3D printers as well as in see-through heads-up displays for automotive navigation and avionics. For Google Glass, the company uses a prism lens that steers the light of images from a pico projector to the lens placed in front of the human eye to view. In this case, the pico projector is designed to be an ultra-short throw projector on the side of a head-mounted display. In 3D printers, they are used to project lights in special wavelengths onto photopolymer materials. It enables a personal stereolithographic apparatus 3D printer, with the object formed additively through a series of light-activated polymerization. Adoption of pico projectors in these markets is low at this time and will take several years to establish.

User Advice:

• For image projection applications in mobile devices, users should pay particular attention to battery life and brightness of projection. Laser light sources are likely to provide the best solution because the image does not require focusing, allowing projection on any uneven surface. These laser versions are, however, much more expensive.
• For image projection applications in wearable electronics, users should wait until Google Glass and other head-mounted displays are more widely adopted by consumers and become affordable.
• 3D printer manufacturers should seek emerging versions of the technology that can project different wavelengths of light. This will allow the combination of multiple photopolymer compounds in order to produce different kinds of objects.

Business Impact: The most obvious business impact is the potential productivity improvements for sales staff who have short periods of face-to-face time with key prospects. The ability to quickly run through a few slides or show a short video can be very beneficial. In addition to phone-embedded pico projectors, the handheld, stand-alone micro projectors are good choices too.

In the longer term, pico projectors embedded in smartglass or indoor furniture that work with sensors may enable spatial and body gesture recognition technology to improve training, safety, customer service and productivity, and have the potential to bring new efficiency to enterprises.

Benefit Rating: Moderate

Market Penetration: Less than 1% of target audience

Maturity: Emerging

Sample Vendors: 3M; Asia Optical; Himax Display; MicroVision; Texas Instruments; Young Optics

Analysis By: Brian Blau

Definition: Virtual reality (VR) provides a computer-generated 3D environment that surrounds a user and responds to that individual's actions in a natural way, usually through immersive head-mounted displays (HMDs) and head tracking. Gloves providing hand tracking and haptic (or touch-sensitive) feedback may be used as well. Room-based systems provide a 3D experience for multiple participants; however, they are more limited in interaction capabilities.

Position and Adoption Speed Justification: Virtual reality is used in high-end simulation and training applications, including military simulation and training like flight simulators, truck operator training in specialized environments (such as mines), and accident investigation in several industries. It is also used in scientific visualization and modeling, including geomodeling in the oil industry and genome mapping, as well as for product design, where VR systems are used to experience automobile or equipment design. Fully immersive theme park rides are also considered VR because of their use of computer-generated graphics, but they are limited to a playback experience and are often not responsive to user input. Entertainment-based VR that uses immersive and interactive storytelling techniques could disrupt major markets by creating new types of entertainment experiences and user interfaces versus the traditional movie theater and television flat screen approach that has been the norm since movies were invented more than 100 years ago. Medical professionals will use VR for telepresence doctoring or even remote surgery. Immersive military applications are more advanced than other types of VR, and the time to plateau of 5 to 10 years is consistent with adoption in the consumer and more traditional consumerlike usage for businesses.

VR experiences are typically used with HMDs. The most well-known is the 2012 crowdsourced Kickstarter project Oculus Rift that raised $2.4 million to develop a fully immersive HMD that is targeted at the video game industry. A retail version of Oculus Rift is still in development and, in an unexpected move, Facebook acquired Oculus in April 2014 for $2 billion. In March 2014, Sony announced Project Morpheus, which is their HMD and VR effort that is targeted toward console gaming. Now that Oculus and Sony are competing for VR experiences, the consumer VR market will likely take off by 2015 and that could usher in a true expansion of the VR technology and solutions market.

Despite these well-established niches and recent HMD advances, reliance on specialized interfaces has kept VR from becoming a mainstream technology in business or entertainment. Virtual worlds such as Second Life and IMVU — which show a 3D environment on a 2D screen rather than immersing the user inside a room or by using an HMD — and the more recent rise of Oculus Rift have all contributed to the rekindled interest in VR. However, major technology and usability advances are still required for a low-cost, broadly used immersive virtual environment. In the meantime, growing popularity of 3D entertainment using 3D glasses — and, increasingly, 3D smart television screens and projections that do not require glasses — may relegate immersive VR to permanent niche status. Currently, augmented reality applications (which superimpose information on the user's view of the real world rather than blocking out the real world) or mixed-reality scenarios (where HMDs and context-aware software are used in a hybrid augmented/virtual environment) are more popular technology approaches to the problem of marrying immersive VR to a consumer setting.

While VR can be amazingly sophisticated, the level of customization can come at a high cost. Recent advances in consumer technologies may help ease these obstacles. Standards (such as for artificial intelligence scripting, object metadata and avatar identity data) are becoming more popular due to the increased use of social networking technologies (such as management of social media profiles) and broadening use of public identities. On the development side, technologies like cloud graphics processing units and mobile video games, as well as the proliferation of broadband access, will allow application developers to integrate VR more easily into their products.

User Advice:

• Evaluate VR for video game development or mission-critical training and simulation activities because it can offer higher degrees of fidelity than simple screen-based systems.
• Consider VR for exploring design issues in the early stages of decision making for high-cost products or architectural designs.

Business Impact: Virtual reality can support a wide variety of simulation and training applications, including rehearsals and response to events. VR can also shorten design cycles through immersive collaboration, and enhance the user interface experience for scientific visualization, education and entertainment.

Benefit Rating: Moderate

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Barco; Digital ArtForms; Facebook; Mechdyne; Presagis; Sony; Virtual Heroes; Vuzix; WorldViz

Analysis By: Stephen Prentice

Definition: A virtual world is an online, networked virtual environment in which participants are immersed in a 3D representation of a virtual space, and interact with other participants and the environment through an avatar (a representation of themselves in the virtual space). This technology profile excludes the extensive virtual worlds incorporated into online computer games.

Position and Adoption Speed Justification: Virtual worlds have moved through the hype phase of development and, while remaining in the trough of the Hype Cycle, are slowly climbing the slope with a limited range of viable application scenarios emerging. The publicity and hype that drove consumer- and social-targeted virtual worlds, such as Second Life, have now dissipated, growth has stagnated, and usage is in decline as Facebook and similar alternatives dominate the online social space, and as video collaboration products emerge with a less intrusive user interface.

The future for virtual worlds appears to be limited to immersive learning environments for education (especially in the commercial sense — training individuals to operate safely in otherwise hazardous environments such as military personnel, emergency services and medical personnel). It is to be expected that concurrent improvements in human-computer interfaces (such as Oculus Rift) and gesture control/movement tracking will enable more intuitive control, compared to the intrusive and overly complex user interfaces of the past, and will make human control of an avatar more intuitive. Compared to the rather dissipated audience value proposition of the past, which was overwhelmed by the inexorable growth of Facebook and similar social sites, the audience for virtual worlds is now much clearer and focused, but much smaller and less exciting, with the emphasis on education and training in immersive environments.

User Advice: The value of virtual worlds to enterprise and educational users resides almost exclusively in the ability to deliver a rich and immersive environment for collaboration and interaction in closed private environments. Current business conditions encourage enterprises to examine a wide range of alternatives to face-to-face meetings and public events, but continuing advances in Web-based videoconferencing provide an easier-to-use alternative for most cases.

Business Impact: The buzz that drove virtual worlds has died, and while their value in "serious games" and immersive environments for simulation and training is now well-established, it is unlikely that additional broader uses will emerge. More intuitive user interfaces may emerge on the back of advances in human-computer interaction technologies, but these will simply enhance existing application cases rather than create new opportunities. Having said that, the acquisition of Oculus Rift by Facebook rekindles the possibility of a large social-oriented virtual environment built on the back of the existing social network. The scale and influence of Facebook means that a second peak of hype is a serious possibility.

Benefit Rating: Low

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: Linden Lab; OpenSimulator; VastPark

Analysis By: Mikako Kitagawa

Definition: 3D LCD technology enables users to view 3D images either with 3D glasses (stereoscopic LCDs) or without 3D glasses (autostereoscopic LCDs). For smaller screen sizes (typically up to 10 inches), directional backlight 3D is a cost-effective technology that uses the LCD's backlight to change the viewing fields for the left and right eyes sequentially, thereby creating image depth.

Position and Adoption Speed Justification: 3D LCD technology has been on the market for some time. However, it remains a niche market aimed at a limited audience of home users (mainly gamers) and select industries, such as aerospace, healthcare and automobile design.

The awareness of 3D technology was highlighted by the success of the movie "Avatar," released in 2009, yet consumers' adoption of and demand for 3D technology has never taken off. The PC-related products with 3D viewing capability are rather limited. The development of "glasses-free" solutions was expected to extend the market for 3D LCD products, as wearing the headgear is one inhibitor to 3D adoption by consumers. Additionally, glasses-free 3D can also be adopted in areas such as advertising and museum exhibitions to deliver a more immersive viewing experience. However, the products of glasses-free development were not well-adopted by consumers, and Toshiba, which introduced a glasses-free 3D laptop in 2012, is no longer manufacturing glasses-free 3D laptops as of 2013.

Almost all major PC vendors once offered at least one 3D-enabled laptop, but only a few vendors still offer 3D-enabled laptops as of 2014. Remaining vendors such as Asus focus more toward gaming devices. The PC industry has lost some of its enthusiasm for 3D, as it could not drive consumer interest, as well as the limited availability of 3D content. One of the fundamental problems of 3D adoption by the mass market is its user experience. Some people feel uncomfortable seeing the 3D vision, so 3D technology will not be a mainstream technology; rather, it remains a niche technology.

For tablets, a few 3D-enabled devices are available, but it is still a very limited market.

User Advice: 3D LCDs remain a niche product while industry hype around the technology has faded.

The technology has uses in both professional and consumer markets; for example:

• In the professional market, 3D technology has been adopted in select industries, such as healthcare and research. Recent drops in the prices of 3D-enabled devices could stimulate adoption in other industries. Professional users should monitor the development of 3D, especially for application development, and assess whether the benefits of this technology justify the price.
• In the consumer market, 3D LCDs are limited to select markets, such as the gaming PC market. However, the 3D-enabled tablet market has potential opportunities beyond gaming, such as mapping.

Business Impact: 3D LCDs are used in some devices, including notebook PCs, PC monitors, games consoles, TVs, mobile phones, GPS units, medical equipment and large-scale design environments. 3D-enabled products are unlikely to achieve mass adoption, but takeup should increase gradually as more products and applications become available. Device vendors need to work closely with application developers to create attractive 3D applications.

PC vendors should also expect takeup of 3D LCDs to remain slow in both the professional and consumer markets, as these products are unlikely to show significant growth, but 3D-enabled PCs could generate good margins and strengthen vendors' claims to be technological leaders.

Industry attention to 3D-enabled tablets has increased, although there are still very few vendors manufacturing them. As a content consumption device that is affordable, a good quality 3D tablet will have opportunities. However, mass adoption will be achieved only with the development of useful 3D content, such as mapping.

Device vendors should look into developing glasses-free 3D technology, as this aspect is one of the issues hindering consumer 3D adoption.

Benefit Rating: Low

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: Asus; Dell; LG; MSI; Nvidia; Samsung

Analysis By: Stephen Prentice

Definition: Ambient and glanceable devices and displays are a distinct category of information appliances that are designed to be integrated into the home and work environments. They convey minimum and specific information in a way that is designed to exploit the "preattentive" processing ability of the human brain. This enables users to absorb the information without having their attention distracted from foreground tasks.

Position and Adoption Speed Justification: Ambient displays are, by their very nature, low-profile, but the basic concepts are sound, supported by extensive behavioral and sociological research. This technology has been moving slowly through the Hype Cycle, but is likely to become very widespread in the long term, and the opportunities arising from the Internet of Things and wearable computing seem likely to accelerate progression. Ambient display devices are defined by these characteristics:

• Simple displays to convey minimal information
• Wirelessly connected to a remote information source
• Not interactive (although they are increasingly incorporated into devices that support interactivity as well)
• Designed to blend into the physical environment
• Designed to convey information to humans without breaking foreground thought

Ambient displays are now being incorporated into a variety of consumer electronic (CE) devices — for example, bedside clocks that change color to indicate changing humidity or pressure combinations that (at a simplistic level) indicate possible weather conditions. In addition, LED home and office lighting technology that is digitally controllable (for brightness and color) is now progressing very quickly in price, performance and availability, and this creates excellent "display infrastructure" for ambient information. As the Internet of Things grows, we expect to see numerous examples emerge. The Good Night Lamp is one early case. It illuminates an Internet-connected lamp when the light in a larger version located elsewhere is activated. It is designed as a simple, nonintrusive indicator of remote presence — a classic application for an ambient display. As the Internet of Things grows, we anticipate numerous similar examples emerging in a steady stream. The new generation of wearable computing devices and the more-complex handheld devices (such as smartphones) are increasingly incorporating ambient-display-like capabilities in "sleep" mode — such as becoming a clock when placed on a surface at negligible additional cost.

As the Internet of Things brings more devices online, the volume of data available demands new forms of visualizations. At the same time, the general environment becomes a more interconnected array of "information artifacts" that communicate with one another, adapt their behavior and increasingly interact to create more contextually relevant information in an easily assimilable form. This technology is destined for broad, albeit low-key, adoption. This may help to expose the benefits of ambient display devices to a growing audience.

User Advice: Simplicity in a complex world is an attractive proposition, especially where enterprises are attempting to induce a feeling of calm, control and goodwill. Ambient displays are the antithesis of today's information-overloaded environment. Enterprises should consider pilot-scale experiments that select a single item of critical information and deliver it to users' environments via an object that blends into their surroundings. Similar devices are now being introduced for home energy monitoring, and they are being adapted by experimenters to reflect a wide variety of other conditions.

Encourage experimentation, and let individuals play with this technology to discover its potential value. Many examples will have a significant social connection. A good example is GlowCap from Vitality, which incorporates a wireless-enabled device into the cap of a prescription medicine bottle. It provides increasingly noticeable alerts (from a flashing light, through audible reminders, telephone calls, and even emails to friends and relatives) when users fail to take their medicines on time. They will provide monthly feedback to summarize performance (including incentives — that is, elements of gamification) and even initiate replacement prescriptions.

Business Impact: In most cases, the business impact will be (at best) moderate. However, the growing volume of data available, combined with the limited ability of individuals to assimilate the details, is forcing information system designers to think far more critically about which information is material. As the trend in business to consumer moves toward a more social alliance between supplier and consumer, ambient displays will play an increasing (but barely noticeable) role in influencing the tempo of decisions. Inside the enterprise, ambient displays will not only indicate the state of the external business environment, but also reflect the changing ambience of the internal social environment.

In many cases, a viable business case can be difficult to create, although, as costs continue to fall, this will become less of an issue. Soft and social factors (such as reassurance about some environmental input or condition), gentle reminders (such as to take a necessary medicine), or background alarm or warning capabilities are more likely to drive acceptance, but they lack the sharply defined quantitative benefits needed for a solid business case. Success will most likely come from integration into the fabric of our technologically enabled and information-rich environment, although it may be that the basic idea becomes incorporated into other devices (such as lighting) in an increasingly computer-controlled living space.

In an almost parallel development, the characteristics of ambient/glanceable displays are increasingly being incorporated into wearable devices (such as Google Glass) to provide information without distracting the wearer from the primary field of view.

Benefit Rating: Moderate

Market Penetration: 1% to 5% of target audience

Maturity: Emerging

Sample Vendors: Ambient Devices; Vitality

Analysis By: Adib Carl Ghubril

Definition: Gaze control brings about computer action by blinking or changing the direction of one's gaze. Gaze control involves determining the angle or position of a user's visual attention, usually through use of cameras, and choosing from a set of available instructions that are mapped to those positions.

Position and Adoption Speed Justification: Samsung has expanded on its elegant battery management feature that turns handset backlights on or off, depending on whether a user's gaze is deemed to be directed toward or away from the screen, by introducing a next-generation feature that does the same for video playback. Tobii Technology's worldwide demonstrations of gaze control for menu navigation on Windows 8, as well as on Android-based devices, are indications that the technology is really taking a positive turn toward being truly useful in various end applications.

Even the more sophisticated gaze control — one requiring a finer realization of gaze — such as the solutions introduced by The Eye Tribe and Atheer Labs, is putting the gaze control technology in the forefront of people's minds. Also, recent announcements and patent filings suggest that Amazon and Apple will follow Samsung's lead on eye tracking in smartphones. Apple's acquisition of PrimeSense (the maker of 3D visual mapping systems and the original Microsoft Kinect) could be a prelude to supporting menu navigation via gaze control, for example, on the iPad — in a manner similar to what Tobii has demonstrated with Windows 8. Meanwhile, The Eye Tribe shipped its first eye-tracking unit in 2014.

Gaze control also has been applied by video game developers, such as Zynga and Blizzard, to design better play theaters, or even to be used as an input control. Companies like Cogisen are now showing solutions that work without infrared illumination, using only the information gathered by the front-facing camera of a mobile device, thus significantly reducing implementation costs.

Gaze control also will get a boost from the proliferation of cameras on computers. Researchers at Microsoft's PixelSense group, for example, are trying to evolve videoconferencing by experimenting with camera systems that view a room through a display that allows for a more realistic representation of the surroundings, one in which participants in a teleconference seem to be looking directly at each other rather than at a camera. We can expect this kind of virtual reality teleconferencing experience within 15 years.

User Advice: Context-aware applications require locational and behavioral cues that could be derived by recognizing a user's gaze. Consequently, incorporating eye-tracking technology is another step in building more-powerful solutions.

Multimodal interfaces, consisting of speech, gaze, and 2D and 3D gestures, will make for more natural interaction between users and machines. They will also allow users hampered by physical challenges to still operate a computer or machine. Applications related to rehabilitation can make use of this.

Ultimately, gaze recognition could be made to effect control by user-generated gaze gestures, like winking to take a picture — which may be supported by retrofitting Google Glass. Yet, users should be cognizant of eye fatigue caused by conscious ocular muscle control.

Business Impact: Gaze control is a fundamental technology for developing machine control solutions in hands-busy medical or industrial applications, such as an operating room or container-inspection depot. It also can be used to enable driverless cars or e-book reading, whereby an application can be made to react to a reader's focusing on a word by, for example, displaying its definition.

Retailers can leverage the technology to track user interest for in-store merchandising. Marketers can leverage gaze recognition in neuromarketing techniques that ascertain how consumers respond to stimuli from product promotions and advertisements.

It is also used as a user interaction technique for severely disabled or paralyzed users. Its impact could be described as "transformational" in enabling those users to communicate (versus the moderate benefit rating for the technology overall).

Gaze control is a key modality in improving computing devices' ability to respond more naturally to users and, in so doing, raises the devices' rate of adoption in their respective market segments. Smarter devices, able to react to subtle changes in users' response by analyzing gaze, are more compelling to users and, thus, more valuable to manufacturers.

Benefit Rating: Moderate

Market Penetration: 20% to 50% of target audience

Maturity: Adolescent

Sample Vendors: Cogisen; EyeTech Digital Systems; EyeTracking; IBM; Seeing Machines; The Eye Tribe; Tobii Technology

Analysis By: Stephen Prentice

Definition: Gesture control is the ability to recognize and interpret movements of the human body to interact with and control a computer system without direct physical contact.

Position and Adoption Speed Justification: The broad proliferation of forward-facing video cameras on devices, from handheld to large wall-mounted, has accelerated gesture recognition up the slope toward mainstream adoption in a broad range of business and personal applications, beyond the early developments in the gaming sector. In multiuser environments or where more accuracy is required, assisted gesture control — which makes use of additional physical objects (such as gloves and wands with inertial sensors) — can be used to enhance the interpretation or resolution of detectable movements. At the same time, alternate sensing technologies are being commercialized (such as the Leap Motion Controller). These offer submillimeter discrimination within a limited, or desktop-sized, zone. The continuing growth of tablets and large-screen smartphones exposes the restrictions imposed by their limited screen size, and the emergence of "hover and swipe" rather than the existing "touch and swipe" may enable a richer set of commands (and a broader range of applications) for these devices. The most recent development is the explosion of wearable devices (for example, fitness bands, smartwatches and augmented reality devices such as Google Glass) in which gestural movements recognized by video or by inertial sensors play a key role in the user interface. Still to come is the likelihood of gesture control playing a key role in controlling autonomous devices (such as robots) in a mixed human/robot workplace environment.

Composite interfaces (combining gesture, movement, facial and voice recognition) can create a rich, immersive and intuitive interface to deliver new capabilities in very competitive environments. Specific business applications are emerging. Several major retailers are now using in-store virtual mirrors, which use gesture control to enable users to select garments and see them superimposed on their bodies. The ability to interact from a distance (and from behind a window) opens up applications in digital signage, banking and other areas. Healthcare applications (in areas such as physiotherapy, fitness and well-being) are emerging, tracking developments in computer gaming. The ability to control devices without physical contact (or while wearing nitrile gloves) has significant benefits in reducing transfer of infectious materials.

In the near-term and midterm future, we anticipate:

• The traditional control paradigms will no longer be appropriate, due to the growing availability of gesture-controlled devices, the rapidly increasing accuracy of these devices and the growing number of devices requiring control, many of which are becoming embedded into the fabric of our environment. Gesture control allows control from the distant "lean-back zone" to the immediate "lean-in zone," and down to the ultimate "wearable" zone. It is looking increasingly significant as a primary interaction paradigm, with the ability to transform the way humans interact with a new generation of computers.
• With mainstream products in the gaming market now well-established, gesture control is moving quickly through the Hype Cycle, and the growing availability of options advances it from "adolescent" to "early mainstream" in terms of maturity.
• While mainstream adoption in gaming is happening quickly, the time to plateau in the enterprise space will be longer, but the changing nature of devices (especially in the consumer area) is forcing the pace of adoption to accelerate.

A market-defined "language" of universally recognized gestures (similar to what has happened in the multitouch area) will emerge to form the base from which more specialized control can be developed.

User Advice: Gesture control is just one element in a collection of technologies (including voice recognition, facial recognition, location awareness, 3D displays and augmented reality) that combine well to reinvent human-computer interaction, especially around wearable devices. Enterprises should:

• Evaluate handheld and camera-based gesture recognition for potential business applications involving controlling screen displays from a distance (the "lean-back" operating zone).
• Evaluate wearable devices to see where they may be employed to enable new modes of interaction.
• Evaluate the emerging generation of desktop-oriented devices, and consider what role they may play in the "lean-in" operating zone.
• Consider how these may be combined with location-based information and augmented-reality displays.

Even the simplest use of gesture, movement or touch can be introduced to existing products (especially in the handheld space) to enhance the user experience.

Business Impact: The ability to interact and control without physical contact frees the user and opens up a range of intuitive interaction opportunities, including the ability to control devices and large screens from a distance. For smaller desktop, handheld and wearable devices, the ability to control the device without physical contact opens up valuable possibilities in a variety of markets, but especially in healthcare applications (where physical contact may result in the transfer of infectious material). Gesture control also benefits the design aesthetics of touch-based devices, allowing users to avoid unsightly fingerprints on their devices.

Benefit Rating: High

Market Penetration: 1% to 5% of target audience

Maturity: Early mainstream

Sample Vendors: Apple; Atheer Labs; eyeSight; Elliptic Labs; GestureTek; Google; Gyration; iNUI Studio; Leap Motion; Microsoft; Nintendo; Oblong; SoftKinetic; Sony

Analysis By: Tuong Huy Nguyen

Definition: Haptics is a tactile or force-feedback technology that leverages a user's sense of touch by applying vibrations, force or motion to the user's fingertips. This stimulation can assist the technology in creating the perception of virtual objects on the device screen. In a broad sense, haptics can be any system that incorporates tactile feedback and vibrates through a sense of touch.

Position and Adoption Speed Justification: Tactile haptic feedback on mobile handsets has traditionally focused on complementing ring tones. Adoption into other applications stagnated until the recent trend in touchscreen displays. Despite this, implementations that emerged from this hype were not compelling enough to drive further interest in the technology. This is due partly to bad implementations and a lack of high-quality haptic routines that complement the touch experience.

Haptic feedback greatly enhances the user experience for touchscreen devices by providing a tactile response to acknowledge user actions. This is especially important in applications such as automotive when users cannot, or even should not, use their eyes. Visually challenged users also benefit from this technology.

Most mobile phones already provide some type of basic touch feedback (such as a low-fidelity haptic solution, typically for phone vibration mode), and a growing number of developers are leveraging this capability. High-fidelity haptics (using piezo and electrostatic solutions) can provide a richer experience, but is significantly more expensive. Tactus Technology has also introduced a complementary tactile solution that uses microfluidics to produce a dynamic, on-demand, physical raised surface the user can touch. Although Disney Research has been investigating the technology for a few years, it reignited some haptic buzz last year with its Electrostatic Vibration solution (formerly known as TelsaTouch). Bristol Interaction and Graphics also has a solution that creates haptic feedback via ultrasound. Finally, a number of companies are also developing integrated solutions such as Texas Instruments' haptics micrcontroller and Synaptics' touch/haptics chip.

Despite the news and technology improvements in haptics, there has been no significant change in the adoption and perception of the technology. Thus, we have not changed its Hype Cycle position.

User Advice: Consumer electronics devices that integrate a touchscreen interface should use haptic feedback as a competitive differentiator. It can be used to enhance the touch feedback loop and add tactile interest to games, as well as supplement audio by simulating low frequency (bass) sounds.

Business Impact: The use of haptics provides a slight enhancement to touchscreen interfaces by improving the user experience, but this improvement is difficult to translate into increased revenue or cost savings. Haptic feedback offers potential improvements in speed and quality for input and interaction with mobile devices, as well as improves the audio experience from smaller, localized speakers in the device.

Benefit Rating: Low

Market Penetration: More than 50% of target audience

Maturity: Adolescent

Sample Vendors: Artificial Muscle; Disney; Immersion; Pacinian; Redux Labs; Senseg; Tactus Technology; University of Bristol

Analysis By: Stephen Prentice

Definition: Large-surface computers are large-screen (typically 40 inches and above) displays that support direct interaction via touch or gesture. They may be horizontal (built into furniture, such as a tabletop) or vertical (wall-mounted or free-standing). They incorporate the elements of multitouch interactions found in handheld devices, but often recognize multiple users, allowing collaborative use.

Position and Adoption Speed Justification: Large-surface computers were initially targeted at the hospitality and retail industries and carried a heavy price premium. In the general market, the growth of large-screen tablets and "table PCs" (such as the Lenovo Horizon) is well underway and has blurred the boundaries of this category, especially at the low end. As a result, the future for large-surface computers at this time appears to be confined to specific vertical-market applications, where their robustness can justify the price point or where a very large size is required (such as exhibition, museum and retail display walls). Following the acquisition of Perceptive Pixel and the release of the Surface notebook, Microsoft appears to have ceded the limited market to others, such as Ideum with their table- and wall-mounted devices. (It should be noted that tablet devices from Microsoft under the Surface model description fall outside this category.)

Improving robustness for large-screen tablets through technologies such as Corning's Gorilla Glass has resulted in widespread usage of these low-cost devices in the public arena. Recent advances in gesture recognition, such as those demonstrated by Leap Motion, open up a range of alternative technology approaches to creating devices in this category.

In the longer term, the integration of large-screen, touch- and gesture-sensitive displays into the building fabric reflects a growing trend toward making computers invisible as discrete objects, but universally accessible — even by inexperienced and casual users. In most cases, the "computer" element within the device is restricted to graphical display and input processing — relying on remote or cloud-based capabilities for processing, storage and content delivery. In the near and medium term, the current range of large surface computers will be restricted to niche applications such as retail, exhibitions and display. In the longer term, devices of this type may become commonplace within the smart home, but will most probably be derived from consumer-oriented devices and be offered at a much lower cost. As a result, long-term growth in large surface computers is considered limited (a "low lander" on the Hype Cycle).

User Advice: Unless the sheer size (40 inches or more of screen display) of large-surface computers is a primary requirement (unlikely outside of specific niche applications), potential users should carefully consider whether low-cost consumer large-screen tablets and crossover table PCs will meet the need at a substantially reduced cost. The justification for large-surface computers as simple transactional kiosks is not compelling, and these devices remain "features," rather than a "must-have" technology. The growth of large-screen consumer handheld devices incorporating multitouch capabilities has also diminished the "wow" factor, which further undermines the business case.

Business Impact: Large-screen display has applicability in a range of physical environments, including restaurant or gaming tables, customer consultation desks, and meeting rooms. However, the growing availability of touchscreens for PCs and tablets is pushing surface computers into specific niche applications centered on hospitality and retail, where their size and robust construction can justify the cost, or in high-profile media applications.

Benefit Rating: Low

Market Penetration: Less than 1% of target audience

Maturity: Early mainstream

Sample Vendors: Eyefactive; Ideum; Mindstorm; Oblong; PQ Labs

Analysis By: Jim Tully; Amy Teng

Definition: Electronic paper refers to several reflective display technologies that do not need a backlight and can be viewed in conditions of good ambient illumination. They use bistable pixels that remain in a particular state after power is removed. For static images, this results in ultralow energy consumption. Most electronic paper technologies involve physical or mechanical movement within the pixel to facilitate a change from light to dark or change color. The performance level achieved is therefore slower than that of other displays such as LCDs.

Position and Adoption Speed Justification: We class several technologies as electronic paper, including:

• Electrophoretic — The most mature electronic paper technology, notably supplied by E Ink. It utilizes organic plastics rather than glass, giving physical characteristics that are surprisingly rugged. The technology is based on pixels composed of charged particles suspended in a fluid. There is much interest in the development of flexible versions of these displays, such as those produced by Polymer Vision (acquired by Wistron) and Plastic Logic. Higher-speed and color versions are appearing, but performance levels fall far short of comparable LCD or organic light-emitting diode (OLED) displays.
• Micro-electromechanical systems (MEMS) — Notably driven by Qualcomm's Mirasol division, these displays employ tiny plates that move when a voltage is applied. In one position, the plate allows light to pass into the pixel and be reflected to the user as a visible pixel; in the other position, the light is absorbed.
• Electrowetting — An electrical charge causes a colored oil to spread across the surface of the pixel. Absence of the charge causes the oil to retreat into a tiny area at the corner of the pixel, producing a black color. The technology was developed mainly by Liquavista (acquired by Samsung Electronics and recently resold to Amazon). The color and speed characteristics of this technology are superior to the other electronic paper technologies previously examined.

Other related technologies use nanochemical changes, rotation of spherical-shaped pixels, and electrochromic effects. The position of electronic paper on the Hype Cycle is an average of these various subtechnologies.

The initial major applications for electronic paper are electronic books (e-books), signage (in retail and roadside applications) and small, information-centric screens such as some types of music players or pedometers. The most visible uses of the technology to date are in the Amazon Kindle e-book products. Other examples include the YotaPhone and Midia InkPhone from Onyx, introduced in 2013 with dual displays (LCD and E Ink), and the Vanguard ID luggage tag with refreshable content via NFC, coupled with wireless power (Powercast) and app.

Low energy consumption is the main driver in most electronic paper applications. Low cost is another catalyst, especially with low-end mobile phones and devices that do not require full-motion video, as well as other applications such as smart labels and electronic tags. Another stimulus is the readability of these displays in bright sunlight, making them ideal for use outdoors. The refresh speed and color support remain limiting factors, and electronic paper is unable to compete with LCDs in these respects. Vendors have experimented with various devices to sidestep these limitations, but are yet to identify the next killer application. Furthermore, the declining cost of LCDs and the increasing attractiveness of OLED displays are challenging a number of application areas for electronic paper, and this is moderating some of the growth opportunities for the technology.

In terms of color capabilities, E Ink displays have improved a little over the past year whereas other technologies appear to have moved backward; Samsung's divestiture of the electrowetting technology and Qualcomm's restructuring of Mirasol have impacted progression of these technologies. We have therefore left the position of the technology unchanged on this year's Hype Cycle.

User Advice: Users and vendors should evaluate this technology in light of their specific business needs — these will vary by the need for color and video, as required. Applications that involve static images (for example, e-books, labels, debit cards, battery-condition monitors, and low-power, battery-operated consumer goods such as radios and watches) are already appropriate for widespread usage, depending primarily on price.

Other applications requiring color and video will probably find LCDs or OLED displays most suitable. Where ultralow energy consumption requirements dominate, electrowetting- or MEMS-based electronic paper should be considered.

Business Impact:

• Use of wireless battery-powered signage is likely to bring significant benefits to some classes of business, most notably in the retail sector, because of the ability to make remote and frequent changes to the displayed content.
• New product classes and new services could also be facilitated with the help of electronic paper.
• The Internet of Things will promote the addition of low-cost embedded technology into large numbers of everyday objects. Low-cost and low-power displays are likely to become an important component of this emerging business opportunity.

Benefit Rating: Moderate

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: E Ink; Magink.com; Ntera; Qualcomm; Samsung Electronics; SiPix Imaging

Analysis By: Tuong Huy Nguyen

Definition: A gyroscope is a device containing a moving mass that facilitates the sensing of position and motion in three-dimensional space (pitch, roll and yaw). Semiconductor micro-electromechanical systems (MEMS) technology shrinks the gyroscope into a small size using silicon components. Traditional gyroscopes typically utilize a spinning mass, while MEMS versions normally use linear motion.

Position and Adoption Speed Justification: The use of MEMS gyroscopes is growing. When used with accelerometers, gyroscopes provide enhanced spatial precision data for numerous applications — improving the user experience for applications involving motion sensing. Multiaxis gyros can also be used as a higher-performance (minimizing noise and increasing accuracy), power-efficient alternative to a multiple-accelerometer solution (detecting rotation, roll, pitch and yaw) or an accelerometer and magnetometer combo solution.

MEMS gyros have applications across a broad range of market segments, including military, industrial, sports and health, and imaging. We expect a majority of the use will be in mobile devices. Adoption will continue to be in higher-end mobile devices with accelerometers and magnetometers. Demand for more accurate spatial positioning, enhanced interactive experiences involving gesture recognition, and improved sensing will drive growth in MEMS gyroscopes. Growth will also be driven by benefits across numerous applications, such as GPS, location, navigation, safety, healthcare, gaming and imaging. Further advances in the technology, such as nine-axis solutions — combining gyros, accelerometers and magnetometers into one, have been developed to optimize on footprint, cost, performance and power consumption — will make this feature increasingly available across consumer electronic devices.

Despite the benefits, a number of factors potentially hinder uptake for MEMS gyroscopes in consumer electronic devices. Cost remains an issue for more price-sensitive applications, such as mobile phones. The full potential/value of MEMS gyros has yet to be achieved. The technology itself is well-established and relatively widespread, but most applications do not take advantage of this feature. Although companies such as InvenSense have offered software development kits (SDKs) to promote the use of and help developers better leverage the gyroscope, developers continue to rely primarily on a combination of the accelerometer, magnetometer and GPS to achieve the necessary location and motion-sensing information.

Although the manufacturing process for MEMS gyros has matured, the maturity of implementation has not progressed over the last year. Therefore, this tech profile remains in the same position as last year.

User Advice: Portable device vendors should use this technology as a competitive differentiator for their higher-end devices, especially those targeting specific market applications, such as GPS and navigation, imaging, health and fitness, and gaming. Vendors can also leverage the gyroscope to improve user interfaces. Similar advice can be applied to consumer electronics vendors that want to include MEMS gyroscopes in their devices. Enterprises and vertical industries such as healthcare should evaluate the benefits of improved accuracy from MEMS gyroscope-enabled devices and compare them with current, as well as alternative, solutions (as MEMS gyroscopes will likely be limited to the high end in the next two years). For less cost- and power-sensitive markets, earlier implementation will be more viable.

Communications service providers (CSPs) and application developers should work together to provide optimized content and services to gyroscope-enabled devices.

Business Impact: MEMS gyroscopes will provide a higher level of accuracy because of the additional data they obtain. Consumer benefits, such as improved user interfaces, a more immersive gaming experience, higher image quality, and more accurate navigation and location, can be leveraged by CSPs, handset vendors and application developers. Improved accuracy can also be leveraged by enterprises for uses such as fleet management and navigation, as well as vertical industries such as medical for healthcare monitoring. In both segments, the gyroscope can be a valuable competitive differentiator.

Benefit Rating: Moderate

Market Penetration: 20% to 50% of target audience

Maturity: Early mainstream

Sample Vendors: Analog Devices; InvenSense; STMicroelectronics

Analysis By: Adib Carl Ghubril

Definition: Bone conduction is the process of propagating sound waves by transferring air vibrations into human bone vibrations. Sound is transferred by either indirectly vibrating the middle ear using a transducer placed against the cheekbone or indirectly vibrating the inner ear using a transducer implanted into the skull.

Position and Adoption Speed Justification: Bone conduction technology was first applied by the military because it allows field soldiers to maintain audio communication without overloading their hearing, which they use to remain cognizant of their surroundings. Even though its audio fidelity is subpar, this technology is finding its way into commercial applications because of the stylistic and design freedoms it enables since it does not require direct access to the eardrum and therefore does not need an uncomfortable earpiece, which often restricts movement and affects personal style.

The first commercial forays were with hearing aid manufacturers that developed invasive systems that were implanted into the skull. Although osteointegration is useful when trying to guide sound waves in a specific manner, it is not necessary for proper functioning. Still, companies such as Sophono are improving the performance of subcutaneous applications by applying techniques that result in indiscernible evidence of surgery and making implants immune to electromagnetic radiation from medical imaging. Sonitus Medical circumvents the surgical process altogether and offers a hearing solution that transmits sound via the wearer's teeth.

Several companies — notably, AfterShokz and Panasonic — have introduced commercial headphones that apply bone conduction on the surface of the skull through headsets that simply wrap around the back of the head. They also introduced new products in 2013 that improve the headset's wearability by providing the user with playback control. Furthermore, the latest bone conduction headsets from Panasonic are equipped with noise cancellation, which should enhance audio quality. Google introduced Glass — augmented reality headware — that utilizes bone conduction for both stylistic and foundational reasons because, by definition, wearable computers are instruments for mediating one's surroundings and must not monopolize one's senses. Max Virtual outfits baseball caps and bicycle helmets with microphones and bone audio transducers to enable hands-free, earbuds-free phone call interaction.

Although various products were introduced, not enough strides were made in signal processing innovation. Bone conduction technology's underwhelming audio experience, which hampers adoption, will necessitate additional signal processing. If improvements remain elusive, this technology will remain in the Trough of Disillusionment.

User Advice: Users may entertain bone conduction hearing systems when uninterrupted concentration on a primary task is paramount. These use cases are typically set in hazardous conditions or in critical situations in which failure has fatal consequences, but they are also applicable in augmented reality applications in which users want to mediate their environment.

Business Impact: Bone conduction technology could raise the total available market of audio systems by being embedded in safety products, such as helmets, whereby a jet fighter pilot, a race car driver or a motorbike driver may remain sensitive to environmental cues, while receiving instruction or listening to music.

Ultimately, bone conduction technology will prove to be a replacement technology to the conventional headset, once audio performance is at par, because of the stylistic freedom it provides.

Benefit Rating: Low

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: AfterShokz; Cochlear; Google; Max Virtual; Panasonic; Sonitus Medical; Sophono

Analysis By: Bern Elliot

Definition: Rich presence is the ability to aggregate and publish presence and location information from multiple sources. Presence can include information such as status, from applications, devices and networks, including device type and release level, as well as from external sources, such as location services. Rich presence is distinct from standard presence, which is usually applied to information from a single source, such as instant messaging (IM) presence.

Position and Adoption Speed Justification: Most leading unified communications (UC) vendors now offer rich presence as an integral part of their UC solutions; however there are both inhibitors and drivers influencing adoption rate.

Adoption drivers are largely due to the wide range of rich presence information that can be leveraged by applications and work processes. For instance, presence services can be leveraged in contact centers. Information about individuals with common skills or responsibilities can be formed into a collective group presence, which can be used to display the availability of anyone in the group.

Presence solutions can also be integrated with applications that aggregate and transform information about context from multiple sources, including community, identity, process and the environment.

One potential area of concern is that presence information about individuals, such as location and availability, may pose real or perceived infringements on privacy. As a result, planners should carefully consider the privacy implications of any deployment. It may be worthwhile to develop a corporate policy for use and to obtain a review of this policy from the corporate legal team.

Inhibitors include organizational, social and technical adoption barriers. As a result, rich presence technology will continue to grow at a moderate rate.

Social barriers include questions about:

• What could be considered personal information
• How information could be used to control workers
• If the information could be used for criminal activity
• If the information could be used to invade personal privacy
• If the information could be used to deliberately mislead others

In addition, some users don't want to change their communication methods, and there are organizational patterns that do not encourage desktop collaboration.

Organizational barriers include concerns about privacy, security, identity and administrator policy. In many cases, these will vary by organization.

Technical barriers largely involve opening up vendor presence services to allow third-party integrations. A proposed Internet Engineering Task Force (IETF) standard will extend the presence states — rich presence extensions to the Presence Information Data Format (PIDF). Vendors compete in some of these areas, rather than agree to a common approach. There are also scaling issues; if not properly managed, then networks can become inundated with presence status change messages. Finally, there are ways to describe similar states that make it difficult for applications and users to share presence status, especially among organizations.

Two standards are used to support presence — XMPP and SIMPLE. Most leading solutions allow integration with either, although XMPP is the more commonly used standard.

An example of presence might be a presence indication of "on the phone," which could aggregate all the user's voice devices: desktop phone, mobile phone, remote office phone and voice over Internet Protocol (VoIP) devices. An alternative example is a virtual meeting, which might have the presence status of "in session," if the following conditions are met:

• A quorum of participants is logged in
• The session chair is logged in
• The recording secretary is logged in

An example of physical equipment presence is mobile defibrillator equipment in a hospital, which has "available" noted. In this example, it is useful to know whether the equipment is available, as well as the location of the equipment.

User Advice: Enterprise planners should investigate how to leverage their presence services to improve employee productivity in structured and unstructured tasks. Presence services could be used to speed workflows and improve business processes. When considering new UC and collaboration investments, planners should review how they will integrate and leverage enterprise or third-party-hosted presence aggregation services.

Business Impact: Rich presence technology can allow enterprises to improve operations in significant ways by integrating context with business processes.

Benefit Rating: Moderate

Market Penetration: 20% to 50% of target audience

Maturity: Early mainstream

Sample Vendors: Alcatel-Lucent; Avaya; Cisco; Google; IBM; Microsoft; Unify

Analysis By: Ant Allan

Definition: Biometric authentication methods use unique traits to verify users' claimed identities when users access endpoint devices; networks; or mobile, networked or Web applications. Across a wide range of use cases, any biometric authentication method may be used in one-to-one comparison mode (when the user enters a user ID to claim a specific identity) or one-to-many search mode (when the user simply presents his or her biometric trait, with no explicit claim of identity, and the system determines his or her user ID from a range of candidates).

Position and Adoption Speed Justification: Biometric authentication methods embrace a variety of discrete technologies, which are used in a wide range of use cases with different demands. The position and time to plateau of this technology represent the optimal cases; individual technologies in particular use cases may be less advanced and may never reach the plateau.

The main drivers for adoption of biometric authentication methods are:

• Improved user experience (UX)
• Increased trust or — particularly — accountability

The potential for improved UX is not always fully realized. Gartner finds that, for some biometric modes, notably fingerprint, at least some users (a few in a thousand) are unable to reliably use this authentication method, and many users (up to 15%) have problems some of the time.

Biometric authentication methods (chiefly fingerprint modes) that are limited by the need for specialist capture devices (sensors) and the cost and complexity of deployments remain niche. Offerings that can exploit the existing inputs on various endpoint devices for voice recognition, face topography, keyboard or gesture dynamics and so on have matured significantly over the past few years. The lower cost and improved convenience of these approaches are steadily driving increased adoption. The trend was underscored by last year's acquisition of PassBan, a mobile biometric authentication specialist, by RSA, the Security Division of EMC, which should make these methods available to many thousands of organizations later this year.

This is particularly relevant to mobile use cases in which traditional authentication methods are falling out of favor because UX is unsatisfactory, the level of trust can be eroded, or technical integration is infeasible or prohibitively costly. UX considerations are particularly important in mobile banking, where clients tell us that there are woefully high login abandonment rates with password-based authentication, and thus they are adopting biometric methods to provide a more transparent or frictionless user authentication.

User Advice: Organizations should evaluate the potential benefits of biometric authentication methods against the needs of each use case and choose among the broad range of biometric technologies on the same basis. Although biometric authentication methods can provide medium to high levels of trust, established, nonbiometric alternatives (such as the several kinds of authentication tokens) are available at a similar price point.

Biometric authentication methods can provide a higher level of individual accountability than alternatives and, so, should be favored when this is paramount. These methods should also be evaluated if UX is a primary consideration, but bear in mind that UX varies across biometric modes. Modes such as face topography and keyboard and gesture dynamics can be implemented in a passive way to maximize UX. Combinations with other authentication methods can increase trust.

Give particular consideration to biometric authentication methods in mobile use cases, where a variety of vendor solutions can exploit the existing inputs on a smartphone or tablet.

Industry hype around the Apple iPhone 5S with Touch ID and the Samsung Galaxy S5 have focused attention on device-embedded biometric authentication, in which capture devices (sensors) and biometric feature extraction, comparison and matching software are built into an endpoint device. Whatever its appeal for consumers, this has limited value for organizations, especially for retail customers and for workforces in companies embracing BYOD practices: Across different endpoint devices, there's uneven availability, which gives rise to inequalities in UX and complicating things for help desks, and where it is available, the value is limited by engineering decisions made by handset and OS vendors, which tend to favor processing efficiency and UX over trust (for example, to reduce false nonmatch or rejection rates). Early implementations (specifically, PayPal's announced support for Samsung Galaxy S5 embedded fingerprint authentication) have yet to demonstrate clear value to the organizations.

Business Impact: Biometric authentication methods can provide improved UX (although this can be inconsistent with some modes) and increased trust and accountability (because biometric traits cannot be easily shared with coworkers as passwords and tokens can).

Biometric authentication methods can provide appropriate levels of trust and accountability for access to corporate networks and Web applications from mobile phones and tablets, where users — especially retail customers, among other external users — will resist having to use any kind of authentication token separate from the device. Biometric modes that offer improved UX are approaching the ideal transparent or frictionless user authentication method that clients tell us they seek for mobile banking and other use cases.

Additionally, some kinds of biometric technologies, such as face topography and keyboard and gesture dynamics, can also mitigate "walk away" risks by providing continuous authentication throughout a session, potentially eliminating the need to enforce timeouts, which users loathe.

Benefit Rating: Moderate

Market Penetration: 5% to 20% of target audience

Maturity: Early mainstream

Sample Vendors: 1010data; AGNITiO; Apple; Auraya Systems; AuthenWare; BehavioSec; BioCatch; BioID; BIO-key; DigitalPersona; EMC (RSA); EyeVerify; Fujitsu; Hitachi; IdentityX; ImageWare; Nuance; Precise Biometrics; Samsung

Analysis By: Angela McIntyre

Definition: Handwriting recognition systems use pattern matching to convert handwritten letters and characters into corresponding computer text or commands in real time. Writing can be with a capacitive stylus or finger on premium smartphones, tablets and PC screens.

Position and Adoption Speed Justification: Handwriting recognition is climbing the Slope of Productivity because it provides a more convenient user interface on smartphones and tablets. Handwriting recognition is slowly climbing the Plateau of Productivity as more people are trying it out. Numerous apps can be downloaded from online stores that leverage the handwriting user interface on different form factors. Handwriting recognition apps are available to download from online stores for note processing, equations and messaging on smartphones, tablets and computers, using OSs, such as Android, iOS, Windows, Symbian and Linux. Professional apps are available for form development, for whiteboards in classrooms, on smart watches, and for handwriting interfaces in automobiles; for example, navigation systems. A leading provider, MyScript, has a software development kit with which partners can build their own custom handwriting recognition-enabled applications. There are numerous handwriting recognition apps available for free on Google Play and the Apple App Store and others can be downloaded for a fee, such as the PhatWare WritePad for iPad for $10 and the MyScript Stylus for $20.

Windows 8 OS has a handwriting recognition feature for touch-enabled PCs in which a software text box accepts handwriting and converts it to text. Microsoft Surface Pro 2 supports handwriting recognition. Handwriting can be entered into Office applications, such as Word, the Notepad application, and any text box in a website opened in Internet Explorer. Editing converted handwriting can be cumbersome in the text box. Mac OS X has a handwriting recognition feature called Inkwell. It enables users to write on the screen using a mouse, or on a peripheral graphics tablet using a stylus. Google Docs has a feature to convert writing into digital text and insert it into a document.

Handwriting recognition has developed for filling out forms, taking notes or writing a few short paragraphs. However, the technology is not mature enough to handle pages of hastily scrawled writing. The technology enables handwriting recognition for printing, script and Asian characters. Written "gestures" in place of tapping icons — for example, for "enter," shortcuts and editing marks — make revising handwritten text faster. Digital pens, such as the LogiPen, can capture handwriting, which can be converted into editable text using MyScript. Furthermore, many of these pens require special paper that adds a high consumable cost to the initial investment in the pen itself, such as the Sky wifi smartpen. In many applications, digital pens will be a stopgap solution that ultimately will be replaced with input into media tablets and other mobile devices.

User Advice: Use handwriting recognition as a convenient alternative to a software or hardware keyboard, for certain use cases:

• Users of Android tablets, iPads, Windows 8 OS tablets and convertible PCs may find handwriting recognition convenient or faster for forms, notes, equations, instant messaging and writing or editing a few short paragraphs.
• For the smartphone, use handwriting recognition for taking notes and instant messaging. A stylus can be convenient for writing Asian characters on a small screen.
• Writing with a capacitive pen instead of a finger can be more convenient when filling in fields on forms or writing more than a few words at a time, especially on devices with screen sizes smaller than 6 inches. Users may find a pen with a coated tip makes their handwriting clearer on devices having slick, high-gloss screens because a "rubberized" tip adds friction between the pen and the glass.

Business Impact: Business use of handwriting recognition includes mobile text entry on tablets or convertible PCs. Handwriting recognition makes a reliable replacement of "clipboards" for forms. Mobile workers in the field and workers in customer-facing roles — for example, sales, inventory management executives, doctors and surveyors — may find using handwriting convenient, especially in the verticals industries of insurance, financial services, healthcare and government. Education will find handwriting recognition useful for converting written notes into editable text.

Benefit Rating: Moderate

Market Penetration: 5% to 20% of target audience

Maturity: Adolescent

Sample Vendors: Digital Field Solutions; Evernote; Microsoft; MyScript; Nokia; Nuance; Paragon Software Group; PhatWare

Analysis By: Adib Carl Ghubril

Definition: Speech recognition systems convert human speech into text or machine instructions.

Position and Adoption Speed Justification: Speech recognition has gained the momentum it needs to move more rapidly toward mainstream adoption as vendors recognize its value in enriching touch and in-air gesture interactions. Speech is a primary form of human interaction and is now deemed crucial in enabling the notion of users doing what is "natural."

With the top cloud service providers — IBM, Microsoft, Google, Amazon, Apple and Samsung — all mobilizing resources in speech recognition systems, the number of applications making use of speech recognition is rising. Apple's purchase of Novauris signals a plan to improve the responsiveness of Siri (Apple's speech recognition engine) by bringing some speech processing back from the cloud and on to the local mobile computing platform. Microsoft, Nuance and others are also tackling dialects and tonal languages.

Dictation, browsing and menu navigation are becoming readily available across PC and mobile platforms. In fact, vendors are now developing systems that recognize dialects in addition to language. Indeed, pattern-matching algorithms have given way to stochastic models (for example, Markov's models) that are now about to be replaced by a hierarchical approach of layered neural networks called "deep neural networks" (DNNs), demonstrating the kind of performance improvement that could bring speech recognition to the required productivity level.

Better noise filtering also has allowed significant improvements in speech recognition in the cabin of the car, and this speech recognition technology is now available in midmarket vehicles.

User Advice: Speech recognition is still very susceptible to the system's immediate surroundings — environmental noise and distance between the user and the microphone dramatically affect performance. Furthermore, cloud-based systems hamper response time, affecting transcription performance and, subsequently, adoption. Indeed, reliable speech recognition will remain elusive until DNN algorithms are fully developed and processing tasks are more effectively partitioned between local processing resources (able to process relatively simple tasks in real time) and virtual resources (able to process relatively more complex tasks but only by introducing lag).

Thus, speech recognition should be deployed "on demand" for individual users who express interest and motivation (for example, those with repetitive stress injuries). Users who are already practiced in dictation will likely be most successful. Also, examine non-mission-critical applications, in which a rough transcription is superior to nothing at all, such as voice mail transcription and searching audio files. In addition, consider speech recognition and its related technology, text to speech, for applications in which users must record notes as they perform detailed visual inspections — for example, radiology, dentistry and manufacturing quality assurance.

For mobile devices, focus initial applications on selecting from lists of predefined items, such as city names, company names or musical artists. This is where speech recognition has the strongest value-add by avoiding scrolling embedded lists while maintaining a high level of accuracy.

Business Impact: Speech recognition for telephony and contact center applications enables enterprises to automate call center functions, such as travel reservations, order status checking, ticketing, stock trading, call routing, directory services, auto attendants and name dialing. Additionally, it is used to enable workers to access and control communications systems, such as telephony, voice mail, email and calendaring applications, using their voices. Mobile workers with hands-busy applications, such as warehousing, can also benefit from speech data entry.

For some users, speech input can provide faster text entry for office, medical and legal dictation, particularly in applications in which speech shortcuts can be used to insert commonly repeated text segments (for example, standard contract clauses).

For mobile devices, applications include name dialing, controlling personal productivity tools, accessing content (such as MP3 files) and using voice-mail-to-text services. Finally, carmakers supporting the control of infotainment and telemetry systems using speech recognition would be addressing a need for unencumbered driving.

Benefit Rating: Moderate

Market Penetration: 20% to 50% of target audience

Maturity: Early mainstream

Sample Vendors: Amazon; Apple; Google; IBM; LumenVox; Microsoft; Nuance; Sensory; Spansion; Telisma

Figure 4. Hype Cycle for Human-Computer Interaction, 2013

Source: Gartner (July 2013)

Table 2. Hype Cycle Phases

Phase	Definition
Innovation Trigger	A breakthrough, public demonstration, product launch or other event generates significant press and industry interest.
Peak of Inflated Expectations	During this phase of overenthusiasm and unrealistic projections, a flurry of well-publicized activity by technology leaders results in some successes, but more failures, as the technology is pushed to its limits. The only enterprises making money are conference organizers and magazine publishers.
Trough of Disillusionment	Because the technology does not live up to its overinflated expectations, it rapidly becomes unfashionable. Media interest wanes, except for a few cautionary tales.
Slope of Enlightenment	Focused experimentation and solid hard work by an increasingly diverse range of organizations lead to a true understanding of the technology's applicability, risks and benefits. Commercial off-the-shelf methodologies and tools ease the development process.
Plateau of Productivity	The real-world benefits of the technology are demonstrated and accepted. Tools and methodologies are increasingly stable as they enter their second and third generations. Growing numbers of organizations feel comfortable with the reduced level of risk; the rapid growth phase of adoption begins. Approximately 20% of the technology's target audience has adopted or is adopting the technology as it enters this phase.
Years to Mainstream Adoption	The time required for the technology to reach the Plateau of Productivity.

Source: Gartner (August 2014)

Table 3. Benefit Ratings

Benefit Rating	Definition
Transformational	Enables new ways of doing business across industries that will result in major shifts in industry dynamics
High	Enables new ways of performing horizontal or vertical processes that will result in significantly increased revenue or cost savings for an enterprise
Moderate	Provides incremental improvements to established processes that will result in increased revenue or cost savings for an enterprise
Low	Slightly improves processes (for example, improved user experience) that will be difficult to translate into increased revenue or cost savings

Source: Gartner (August 2014)

Table 4. Maturity Levels

Maturity Level	Status	Products/Vendors
Embryonic	In labs	None
Emerging	Commercialization by vendors Pilots and deployments by industry leaders	First generation High price Much customization
Adolescent	Maturing technology capabilities and process understanding Uptake beyond early adopters	Second generation Less customization
Early mainstream	Proven technology Vendors, technology and adoption rapidly evolving	Third generation More out of box Methodologies
Mature mainstream	Robust technology Not much evolution in vendors or technology	Several dominant vendors
Legacy	Not appropriate for new developments Cost of migration constrains replacement	Maintenance revenue focus
Obsolete	Rarely used	Used/resale market only

Source: Gartner (August 2014)

Source: Gartner Research Note G00264133, Adib Carl Ghubril, Stephen Prentice, 13 August 2014