Three key players in factory automation team up with Japanese AI venture PFN to develop FIELD Systems, an "edge and fog computing" system that combines CNC, robots, devices and sensors into a self-learning system.
On 18 April 2016, Fanuc announced its Fanuc Intelligent Edge Link and Drive (FIELD) System — developed jointly with Cisco, Rockwell Automation and industry startup Preferred Networks (PFN) — for bringing "edge and fog computing" into the factory environment.
The new FIELD System represents Fanuc’s latest move to bring digital business solutions closer to the factory floor by creating a new partner ecosystem that can provide end users with an integrated manufacturing solution for interconnecting and analyzing data from computer numerically controlled (CNC) robots and sensors in factory environment. This is more than just a conceptual announcement. It evolved from several other related proof-of-concept activities. For example:
Fanuc America and Cisco conducted an 18-month Zero Downtime Trial with a major automotive manufacturer, reducing unplanned factory downtime (estimated at $20,000 per minute) and improving operating efficiency using the Cisco Internet of Things (IoT) cloud.
Fanuc and Rockwell Automation previously offered a pre-engineered integrated automation solution that integrates Fanuc CNCs and robots with Rockwell Automation cell controllers.
What makes the FIELD system unique is the “ edge and fog computing” concept , and the addition of deep learning technology from PFN. FIELD not only enables an on-premises near field to handle information from large numbers of devices (rather than connecting devices to an external cloud directly) but also introduces “distributed cooperative deep learning” developed by PFN. This technology enables learning through collaboration of multiple connected machines rather than just a single machine, shortening learning time. Applying this technology into Fanuc robots will accelerate practical use of robots that learn jobs automatically. During a d emonstration of autonomous learning of a bin-picking robot In December 2015, the robot achieved human level performance within eight hours.
The FIELD system operates on Cisco Unified Computing System (UCS), running general applications such as simulation as well as the new deep-learning application from PFN on Windows and Linux OS. Fanuc is trying to create an ecosystem to expand opportunities to integrate factory automation — including robots, CNC, sensors and networks — to propel application development from new partners. Fanuc indicated it is in touch with a number of candidate partners to be disclosed soon.
This is a good example of moving digital business forward — not only for manufacture companies but also for all industry. Fanuc has chosen to invest in deep learning as a key technology to expand its potential. Further, Fanuc has decided to coordinate with Cisco and Rockwell to lead industrial Internet development globally. This approach offers a good mix of current confidence and future investment potential.
CEOs, CIOs and IT leaders in manufacturing:
Invest time and resources into deepening your organization’s understanding of the use of the Internet of Things, robots and artificial intelligence. Digital business is about to become “real” in the factory.
Use this announcement as a trigger to analyze your own manufacturing operations (as well as those of your suppliers) for opportunities to transform automation more widely. Distributed cooperative deep learning technology will open avenues for new digital business opportunities.
Invest in new technologies as they become available. Run trials to encourage “cutting-edge thinking” and development of new skills within your own organization.
Initiate proof of concept projects for some of your most impactful technologies with vendors who have the capability to develop evaluation systems.