Dirk Slama: Testbeds are a key element of the Industrial Internet Consortium strategy. What exactly is a testbed? 

Michael Lee: At its simplest, a testbed is a controlled experimentation platform where applications can be deployed and tested in an environment that resembles real-world conditions. The approximation of, but insulation from, the real world allows the issues of safety, security, reproducibility, etc. to be managed during an experiment. The experimentation itself can explore untested technologies or existing technologies working together in an untested manner. And, most importantly for the Industrial Internet Consortium, testbeds should result in the creation of new products, services, and technological innovations. The Track & Trace testbed, for example, should result in open tool interfaces and proven localization technologies that drive innovation on the factory floor and create opportunities for new products and services.

Dirk Slama: How does a project qualify to become a testbed? 

Michael Lee: For an Industrial Internet Consortium testbed, it has to be approved by the Industrial Internet Consortium Steering Committee. Their overriding approval criteria are the perceived degree to which the testbed can successfully lead to innovative and new products, and how it can verify or extend our technical capabilities. To assess these criteria, numerous characteristics of the testbed are considered, including business case, technology requirements, social and cultural impacts, research requirements, timeline, etc.

Dirk Slama: What type of testbeds are you looking for, and how many?

Michael Lee: Tough question. I’m not sure we know enough to set a limit on either the type or the number of testbeds at this point. The Internet of Industrial Things is here. And member companies of the Industrial Internet Consortium are already driving many usages in industry. It’s a massive enterprise and it will be decades before we know exactly where it may lead. Sir Tim Berners-Lee didn’t predict all of today’s uses of the Internet when he invented the World Wide Web in 1989, and I don’t think we can predict all the uses of the Industrial Internet, or how many testbeds will be required going forward. Our current focus is therefore more on qualifying and approving Industrial Internet Consortium testbeds rather than on limiting their number.

Dirk Slama: The Industrial Internet Consortium says it does not do any standardization work. What does this mean for a testbed? 

Michael Lee: That’s correct, the Industrial Internet Consortium does not create standards. However, standards are important to us, and it is an area we are very involved in. To make this a little clearer, let me briefly explain the Industrial Internet Consortium’s major activities and how standards fit into them. Broadly speaking, there are three main activities, as shown in the figure below.

The first element (on the left of the figure) relates to the ecosystem in general and membership in particular. The business needs from the ecosystem drive ongoing work in the security and technology groups, and all of that leads to the creation of innovative products and services by means of the Industrial Internet Consortium testbed. The security and technology references in the middle element are either based on existing, hopefully open, standards or help to identify a gap or lack of standards. In the latter case, the Industrial Internet Consortium will work to clarify the requirements for missing standards and assist members in standardization efforts along with appropriate (external) organizations. Likewise, the testbed activities may help refine these requirements and possibly vet a potential technology for standardization. This is reflected in the lessons learned-improvements feedback loop above.

Dirk Slama: What does the current testbed pipeline look like, and what’s next? 

Michael Lee: The testbeds are member-driven and thus reflect the needs and expectations of the consortium’s members. In practice, they are typically driven by multiple, collaborating members. There are currently a variety of testbeds in various stages of consideration. While it is too early to comment on specifics, I can comment on some patterns we’re seeing. One is “Horizontal” testbed initiatives. These tend to be technology driven and focus on things like software-defined networks, cross-domain connectivity, and ultra-high-speed networking. There are also “Vertical” testbed initiatives. These tend to be market driven and focus on things like remote patient healthcare and smart tools on the factory floor (as the Track & Trace testbed is doing). And finally, there are “Grand Challenge” initiatives, which have a much broader and longer-term focus and may drive the need for multiple testbeds. These are a popular topic in the IoT world and include subjects like healthcare, transportation, energy, smart cities, etc. Additional effort is required to maintain a sustained and coordinated approach when working on this kind of initiative. I expect we will soon see new Industrial Internet Consortium testbed initiatives reflecting all three of these patterns.