A patient with heart disease who has been instructed to weigh himself daily on a smart scale gets a call from his physician’s office. The call is a preventative check-in, triggered by the patient’s recent, sharp weight gain — which, for people with heart disease, can be a sign of congestive heart failure.
This is an example of the Internet of Things, where devices collect, sense and analyze data to trigger actions defined by business goals. The IoT can transform core business processes (and in this instance, prevent an acute medical emergency or even save a life). It’s paved the way for new business models in areas including predictive maintenance, connected vehicles and smart appliances.
However, cautioned Benoit Lheureux, research vice president at Gartner, “IoT project implementers often underestimate the complexity of IoT integration and overestimate the built-in integration capabilities of their IoT platforms.”
In a Gartner Research Circle survey on IoT, companies implementing IoT ranked integration as one of their top three technical challenges (in addition to cybersecurity and determining and managing business requirements). Mr. Lheureux said, “Gartner predicts that through 2018, half the cost of implementing IoT solutions will be spent integrating various IoT components with each other and back-end systems.”
Three questions to ask about IoT integrations
In the early stages of IoT planning, it’s important to ensure that all key project stakeholders are on the same page, advised Mr. Lheureux. Hold brainstorming sessions to identify initial IoT integration requirements. Three questions should be considered:
1. What are our fundamental integration challenges?
IoT projects typically involve three types of integration challenges: device, data, and applications and processes.
In the smart scale example, the IoT platform must identify, connect to and ingest data — securely and reliably — when the patient steps on the device. (Multiply that by thousands of devices, or even millions in other IoT scenarios). To work, large volumes of IoT device master data must be correlated with existing patient data. And then, new IoT rules need to trigger events, such as scheduling a nurse’s visit, in the existing healthcare practice management software.
2. What complications might arise if the scope of integration expands?
Projects become more challenging when additional levels of IoT integration for back-end applications, cloud-based IT assets (e.g., software as a service), mobile apps and B2B ecosystems are needed. For example, if an organization outsourced smart-scale monitoring to an external provider, it would need to integrate that provider’s data and processes with its own IT assets.
3. How does IoT tax our existing integration skills?
The massive volume of devices involved in many IoT projects far exceeds the number of endpoints that organizations are accustomed to integrating into their virtual enterprises. In the telemedicine example, smart scales and other monitoring devices are a new form of master data added to the mix of sources from which data is collected.
Similarly, both large, complex things (such as buildings) and large numbers of simple things (like appliances) can produce constant streams of IoT data. This creates a need for new data integration capabilities, as well as information management skill sets and resources.
Start small and frequently re-assess
Limit the scope, scale and complexity of initial IoT projects, advised Mr. Lheureux. Utilize built-in IoT integration capabilities at first, then extend those with best-of-breed integration solutions, e.g., integration platform as a service (iPaaS), later. “This gives you an opportunity to quickly capitalize on initial IoT projects, then incrementally utilize new integration skills and technologies to address your expanding integration requirements later.”