Executive Interview: Richard Soley, CEO of OMG


Industrial Internet Consortium Members

“Invent the Future” by Anticipating Disruptions from AI,

Pursue Standard Testbeds to Accelerate Adoption

Dr. Richard Soley is Chairman and CEO of the Object Management Group (OMG) and Executive Director of the Industrial Internet Consortium (IIC). In a career in the computer industry approaching 40 years, he has overseen a range of software standard collaboration efforts. These include the CORBA specification, the Unified Modeling Language (UML) and Model Driven Architecture (MDA), which permeate critical software today. He began his professional career at Honeywell Computer Systems working on the Multics operating systems. He was later co-founder and CEO of A.I. Architects, maker of the 386 HummingBoard. Dr. Soley holds bachelor’s, master’s and doctoral degrees in Computer Science and Engineering from MIT. He recently spoke with AI Trends Editor John P. Desmond.

Q. Please describe the mission of the IIC today and give us your historical perspective on AI?

A. The IIC mission continues to focus on industrial applications and market verticals including smart cities, transportation, and agriculture among others. What I find interesting is this sudden recognition that artificial intelligence (AI), machine learning, and even 3D printing, all help in the adoption of IoT and industrial IoT systems.

I was CEO of AI Architects in the mid to late ’80s and worked for Symbolics, which was an AI company making expert system tools and AI hardware. This was around the same time when “Time” magazine announced its 1984 Man of the Year award as the computer, which I believe generated the AI Winter. The magazine created the expectation that machines think. I remember computer companies like Thinking Machines Corporation, whose motto was, “We want to make a computer that’s proud of us,” and that was never the function of AI. Expectations were way over-hyped and couldn’t possibly be met.

The ubiquitous access today to a large number of connected cloud computers that have far more compute power, far more memory, and access to open source software for data analysis, that’s something that we couldn’t even dream of in 1982. It created an opportunity to put more intelligence in systems and bring us to where we are today. It remains overhyped, because that’s the nature of the IT industry, but it will deliver a lot more. Voice recognition experienced this struggle in 1982. And now, we all carry a voice recognition system in our pockets, whether we call it Siri or Google Assistant.

Q. Can you share how IIC is organized and funded?

A. The Industrial Internet Consortium (IIC) is a major part of the Object Management Group, which is, itself, nearly a 30-year old standards organization. The IIC is accelerating the adoption of industrial IoT by building testbeds. That requires agreement on a shared architecture, agreed security framework, analytics framework, vocabulary, and so forth.

The test beds are IIC’s major differentiator. The largest funding source for IIC is membership with the balance coming mostly from events. Funding of the IIC parallels the funding of OMG.

Q. What are some of the most important IIC initiatives today?

A. The most important initiatives are the test beds run by our nearly 300 member companies. Roughly 30 testbeds are currently running and I’ll give examples of a few compelling ones.

The Track and Trace testbed was initiated by Bosch Software Innovations in Germany. It started with a very simple idea that factories could be made more efficient and safer if you know where everything in the factory was located within a meter. The 3-year old testbed utilizes Cisco Wi-Fi routers to triangulate position of things – people, parts, works-in-progress, and tools – inside the factory and overhead cameras to provide about five centimeters of resolution. Results from this testbed started publishing last year, which are informing the requirements for new standards, new concepts of training, retraining, and hiring. It’s a great opportunity to learn more about IoT in manufacturing.

But we’re not limited to manufacturing. Our Infinite Testbed is managed by Dell Technologies in Southern Island, County Cork, Ireland. In over two and a half years of use, they have integrated national and provincial information resources to optimize ambulance delivery and information delivered to and from ambulances to assist in saving lives.

The Smart Building Testbed is a project between Dell and Toshiba in Yokohama, Japan. The companies outfitted a brand-new building with 35,000 sensors collecting between 0.5-1.0 terabyte of data every day. The sensor data collects information about light, temperature, people movement, telephone calls, and more. And it’s learning how the building is used, which enables optimization for the comfort of the users and predictive maintenance of the building.

Major testbed results are being published in the IIC “Journal of Innovation.”

Q. You’ve been involved in several collaborative efforts to define computing standards, including CORBA, Unified Modeling Language and Model-Driven Architecture. How effective has the effort to collaborate on software standards been?

A. Very effective. There are about five billion corporate systems running today, including every smartphone in the world, every telco switch, every banking system, every robotic system, and so forth. Our oldest standard, now more than 28 years old, is the corporate standard Common Object Request Broker Architecture (CORBA), built into every Java virtual machine.

According to Gartner, 71% of all software development organizations use UML today. Our model-driven approach to building software has been extremely effective.

We’re now in about two dozen vertical markets. And OMG standards drive every retail point of sale, every NATO military radio, and every middleware system. And we have new standards coming out in CubeSats, (a miniature model for space research). We’ve run our standards process over a thousand times, and all those standards are implemented. Implementation is a requirement of our standards process.

Q.Could you contrast the Industrial IoT opportunity with its Consumer IoT counterpart?

A. First of all, there’s plenty of good work going on in the consumer space and IIC does not need to get involved in it. We see bigger opportunities in the industrial space, where IoT can have huge, disruptive effects on markets including agriculture, healthcare, transportation, smart cities, manufacturing, and production. We’re trying to learn what that disruption will be.

And, in terms of the technology, standards and security are mission-critical in the industrial space. If you hit a switch and nothing goes on or off in your house, you just hit the switch again. That’s not the end of the world. But if the factory stops working for a couple of days, that might be the business equivalent to the end of the world.

Q. How does IIC differentiate the testbeds from a use case?

A. A use case is a use of technology. One of the limitations with standards and with test beds is a focus on the technology, instead of the application of that technology. Some of the testbeds that we’ve developed at the IIC, such as the Time Sensitive Network (TSN) Testbed, are necessarily focused on technology availability, integration, and portability.

The use case approach is driven by understanding the desired outcome as opposed to what technology is currently available. For example, the world’s largest copper mining company had a unique need for high-reliability wireless networks to make mining operations safer. As a result, we’re putting together a stack of technology from our members to deliver industry-specific requirements.

Q. Looking ahead, what does the intersection of AI and industrial IoT suggest for innovation?

A. Without question, AI and IoT are going to disrupt enormous existing markets, including transportation, manufacturing, and healthcare. With access to more intelligence, the huge amounts of data generated by industrial IoT can actually be analyzed, generating insights that lead us to better efficiencies, better productivity, and improved safety.

When you have systems that can ingest massive amounts of unstructured data and generate new insights, you are outperforming the human capacity. So, we’re going to see huge disruptions driven by the combination of AI and IoT.

For more information, go to the Industrial Internet Consortium.