Published on 07/09/2018 | Technology
*This episode of the Industrial IoT Spotlight Podcast is sponsored by the Industrial Internet Consortium
What are the only five questions you should ask to determine which communication standard to use? How should we handle the challenge of defining standards? What does the vastness and confusion of the IoT market mean for technology solution providers?
Stan explains to us how the IIC is helping to define standards in communications by clarifying vocabulary. He also explains to us the origin of the IoT hype and how both industrial and technology companies should do to succeed in this market.
The next IIC meeting will be May 25thin Finland, and the Q3 meeting will be in Chicago the week of September 12th to 15th, the Q4 will be in China on November 16. Non-IIC members are welcome to stop by and meet the community so please do make an effort.
IIC Connect: https://www.iiconsortium.org/iic-connect.htm
IIC Journal of Innovation: https://www.iiconsortium.org/news/joi-articles/2017-Sept-Practical_and_Theoretical_Guide_to_Using_IICF.pdf
Stan Schneider is a leader in the Industrial Internet of Things (IIoT), otherwise known as smart machines. He is CEO of the leading connectivity company in the industry, Vice Chair of the largest consortium in the industry, an advisory board member at the industry's largest event, and ranked among the top 10 IIoT influencers worldwide. RTI has experience working with more than 1000 applications across Automotive, Transportation, Energy, Medical, Defense, and Industrial Control. Stan specializes in disruptive innovations caused by the combination of artificial intelligence and pervasive networking.
LinkedIn: Stan Schneider
Welcome back to the Industrial IoT Spotlight. I'm joined today by Stan Schneider. Stan is CEO of RTI as well as vice chair of the board at the Industrial Internet Consortium (IIC). Today we're discussing a couple of topics with Stan. First the ecosystem Task Group which is Task Group that Stan is running at the IIC and the second is the connectivity framework. Stan, thanks so much for taking the time to speak with us today.
No problem. Great to talk to you again. Always wonderful to be involved with IoT One I think you guys are one of the most clue-ful analysts out there.
We try. I think we have a long way to go but we do try and that's the first step. Before we get into what you do at the IIC, can you give us a quick background on RTI? The company has had a really interesting history - you have ridden a few waves of technology in the lifetime of the company. How did you end up as the business you are today?
We started out as a spinoff of a lab at Stanford called the aerospace robotics lab, that is actually now called the Autonomous Systems Lab, that focuses on autonomous, mostly flying, things but also space robots. For a while, we are trying to sell robotics, but there wasn't much of a robotics market. Very long story short the tools we had were very valuable and we became a tools company for a while, and that company did quite well. We sold that company in 2005.
We are experts in autonomy; we didn't have our tools business anymore but we did have a pretty good bunch of people and we got pretty good exit. We looked around and saw that there's a really interesting trend out there device networking. At that time, we called it pervasive information infrastructure for devices, now called the Internet of Things. We decided to go off and build the architecture now called the framework for connecting the physics speed devices to scale the arrival of the Internet, and we're still working towards that vision.
We sell a connectivity framework which is so I'll explain that later. It's a way of putting together devices, intelligent modules, cloud components, and all the things in the Internet of Things, to build more autonomy in the system or in many more kinds of systems.
We're chiefly known to be active in autonomous vehicles, but we are also active in autonomous power systems, medical system, different kinds of transportation systems, and even still some space stuff like launch control systems spacecraft. We live at the intersection of artificial intelligence and pervasive networking. We're not in the AI company; we don't do back proposition or propagation or deep learning or anything like that. We are also not a networking company - we don't do what most people consider networking, and connecting things together with wires and stuff. But we do allow you to take devices and connect them to intelligent modules and build smart stuff from them.
Since the IIC was founded, RTI has been one of the very early members and always very active. Every meeting that I go to, you're there or a couple of people from your team are there. Why do you invest so many resources from yourself and your team in the IIC? What's the value that you see in being part of this community, as opposed to just plugging away building technology and selling?
We very much believe in the Internet of Things, especially in what we call the edge autonomy branch of IoT. I think it really is the future of technology. The thing that is holding it back right now, from our point of view, is just confusion. People don't understand what technologies do it; they don’t understand what opportunities are for there for their businesses in this market. They do see the obstruction. There's a lot of fear out there. But there's not a lot of understanding how you mitigate that fear and turn it into actually for the future.
The IIC came along and really created a market. I give it full credit for creating the IoT market. The Stanford business school definition of a market is a group of consumers with similar needs who communicate and exchange. Back before the IIC, we had lots of projects sort of independent projects that were not very well defined together in many different markets. The IIC was a first place with the horizontal breadth and the technical depths that you could talk about how a controller works for a CT scanner versus a vehicle versus a factory automation system versus a power system renewable plant.
That really caused a place where people could meet and discuss all these kind of things together. It's an ecosystem first, and we got excited about this, probably fits our vision and we should go invest in this so we can help resolve that confusion and build the market. This is also why I started the ecosystem task group.
Why don't we cover that one before we get into the framework? One of the things I really like about your ecosystem task group is that you really make an effort get the smaller companies involved. Naturally a lot of people probably think that the Industrial Internet market is naturally going to be dominated by some of the very large technology companies, because they have this legacy infrastructure in place and they have the R&D to really go deep into technology stacks and allow them to already connect the systems. Whereas smaller players have to figure out where does their technology plug into a complicated system and make an effort to get involved in real business activities outside of the research lab. Can you give us a little bit of background on how did this get kicked off the ecosystem?
In the ecosystem task group, there is a good mix of organizations – small companies, big companies, governments, non-profits, etc. Most of the organizations joining us have a pretty long term view point --they're trying to create standards and technologies on which they can build multibillion dollar businesses down the line, five to ten years out.
Small companies obviously can't work on that kind of time scale. As the steering committee representative for small companies, I was trying to find a way for small companies to leverage the IIC into a couple different directions: partnership and awareness. IIC is a very rich ecosystem; RTI has done lots of great partnerships there. We wanted to find a way to make that more accessible for smaller companies. And that was really the genesis of it: to create an organization whose main purpose is to allow the members to monetize their investments in a in a couple of years, rather than five or six or more years.
The IIC is really doing three things.
First, building an ecosystem where organizations can cooperate and compete to build competitive positions and build partnerships and go to market plans with other companies. The second thing is they put together guidance - how do you actually build your systems with all the technical experts. The last thing is testbed to prove that the guides and the system work.
The ecosystem task group is just a matching service. We hold a session called IIC connected face to face meetings where you can sign up for slots to talk to other companies. It's a great place to spend 20 minutes really digging into what somebody else does. If you're a small company you can try and find partners are used as a technology, and if you're a large company you get to know better what's going on.
Test beds are what technology trends are. So that's that's one service.
The other thing we're doing which will be soon released is IIC ecosystem directory, which is the browsable index of the IIC database.
If you have Google, you can search for all things in the Internet of Things, but you have to know what to search for, and it’s very difficult for you to know that. Just in the connectivity world where I play, there are so many conflicting terms that you never really figure out what a connectivity framework is, or what different standards really do. So we are putting together a directory of the entire market, of the entire ecosystem. When that goes live, people coming into the website should be able to figure out much quicker how the market is structured, what are the technologies available. For example, there are 400+ IoT platforms in the market and they all do very different things. With this searchable database, you can figure out what categories there are, what is the compatibility, etc.
There was a white paper report out recently called “Pilot Purgatory” around the IoT space and a lot of projects that were getting done as pilot projects on a relatively small scale and were for various reasons not ramping up. Some of the reasons are directly related to connectivity and there are also challenges that companies have in figuring out how to scale connectivity. Where do you see this market today in terms of maturity moving out of the earlier stage exploratory projects pilot projects or small scale kind of local projects, towards really mass adoption of systems that are radically changing the businesses that are investing in them?
First we need to talk about what the IoT is. It is three main categories of things.
The first is device monitoring, which includes most of the consumer side of IoT. It consists of a single device that talks to a single cloud service or IoT platform, without involvement with other devices or platforms. This includes Fitbits, thermostats, intelligent doors and lights. Whereas the industrial IoT has use cases like predictive maintenance, optimization of parts, which are simple from a connectivity point of view.
The second is analytic optimization, in which we take a plant of sensors and collect data in the cloud to analyze it and optimize the plant. It is quite a different use case from device monitoring.
The third thing, is edge autonomy, where the real action is not between the devices and the cloud, but between devices and the edge. The cloud may be part of that, but it is about something that is smarter and more active in the field.
There is pilot purgatory in every industry, as the Gartner Hype Cycle illustrates. Everybody thinks that the next technology will solve everything instantaneously, and then when they find out that there is a lot of work to be done to make that happen, they fall into the trough of disillusionment. I think there is definitely a lot of that in the IoT space, especially in the consumer IoT space. People see that as negative, but I see that as an absolutely necessary step to get to the next phase of the market, and an excellent way to define new directions.
RTI has quite a bit of traction mostly in things that have some very compelling reasons to get better, that often has nothing to do with the Internet of Things. For example, autonomous cars, but also renewable energy, renewable integration into the grid. Some things don’t work well today because it is not smart enough, but these are the things going after real problems, not just optimizing something to make it a little bit better.
It is expensive to build new systems and get the technology to a place where it can make real impact. To get there, there will be mistakes, and these mistakes cost money. But they fundamentally change the way the world and the market works. For example, in autonomous vehicles, obviously many of the pilots will not get through the pilot purgatory stage, but it is quite compelling to think that we will eventually get through the pilot stage into a real and sustainable industry.
You’re right that pilot purgatory is something natural in new technologies and not necessarily a bad thing, but just a matter of the technology working its way through development and learning cycles. In this market, it might very well be the connectivity framework – connectivity is one of the areas that needs to be refined in order to make some of these more mass market options feasible. A lot of solutions look good on paper, but do not meet expectations when deployed because of the challenges in moving data between the engine and the systems on a very diverse set of legacy environments out in the real world.
Therefore, finding a way to standardize this so that information can move around cost effectively and reliably is a very high priority. What is the background behind your connectivity framework? What was the impetus behind developing it?
Our first goal was to analyze and understand all the standards out there. I wrote a paper that set out how we were going to plot the usage requirements against what the standards can do. You could have 10 different dimensions of requirements, and when you’re done with that paper, you could understand which standards could solve your problem, and pick one standard based on other factors.
We spent three years getting together companies to analyze the different standards that they use, and came to the shocking conclusion that people are still coming to grips with what is available out there. The standards space is unbelievably immense, and use cases are very big. The standards are so different that there is almost no overlap.
After going through this, I came to the conclusion that if you are in business and trying to build something, you should really spent a lot of time understanding the standards before trying to choose one. If you just choose something that you’re familiar with, you’re probably going to go in the wrong direction. You really need to understand what the various standards do differently, and the differences are so big that with a few questions about any application, we can figure out which standard is used in almost no time. Because they are so far apart, we ended up putting together an architecture for the eventual real industrial Internet of Things, because it eventually needs to be able to integrate multiple standards, now called the core connectivity architecture.
The first challenge would be understanding the correct standard for your application, which is a surprisingly difficult problem to solve – a good amount of research is not necessarily going to solve it. The main problem today is how to deploy a solution across an environment that uses many different standards. For example, when you look at the factory floor with lots of sensors. The fundamental bottleneck in deploying a straightforward IoT solution, in many cases, is simply getting the data from the sensors to the platform. I think this is incredibly important but it is not easy because there are technical challenges due to vertical silos. How are you addressing this issue from the IIC perspective?
From the IIC perspective, we are trying to gain some clarity to navigate the claims out there, and write some vocabulary.
IoT communication protocols consists of 4 factors - publish, subscribe, things, and real-time. They are vastly different because of the way they define each of these terms. Some protocols define “real-time” has milliseconds, and some define it as 24 hours. Some protocols define a “thing” as a thermostat, while others define it as an entire grid. The different definitions overload the terms so that it becomes impossible to navigate. The first thing the IIC did is to put together a framework to analyze different connectivity technology to understand what they can do that is different from each other. RTI is very active because it is in our interest not to try and sell our product in the applications where it doesn’t really make sense.
The Internet of Things is capable of changing the way industries work fundamentally. So the IIC is trying to help all these industries map their applications to the technologies that can make it work, and making these technologies all work together in the future.
From RTI’s perspective, what is your primary value in doing this? Is it to have an understanding of where your solutions would fit into the greater ecosystem and communicating that out for potential customers or partners? Or accelerating your sales process because your customers and partners could be more educated in the technologies?
The framework has already helped RTI a lot because the framework helps us to qualify leads more effectively by understanding customers’ current operating environments. We ask five questions to understand which standards they use and whether they are compatible with our problems, which you can find in the IIC Journal of Innovation. So with this, we can take some confusion out of the industry to help companies find a clear niche where they can really grow and excel.
That would be an excellent online tool because asking somebody to read through a white paper is a heavy ask, but it is much easier to go to a website and answer 10 questions to get a diagnosis of the the possible solutions for your solution.
Most of the questions are quite simple.
Do you believe that gateways are going to the default solution when you are working in an environment with multiple standards, and you can push everything through a standardized gateway and get it up to the cloud?
We have two different types of gateways.
The first is gateways are controlled by a clearly written standard that translates between data models and data systems on either side of the gateway in a known way that allows you to build an interoperable Internet.
The second type is a gateway that connects one of your standards into one of the core standards, and now you should be able to build an entire architecture, without having incredibly powerful gateways that can handle many different kinds of standards.
That’s where we are today. In the future, there will be more and more core standards that allow you to connect different subsystems in a larger network. There will always be some use of taking unusual standards and connecting new or unusual protocols which have nothing to do with a standard, or connecting legacy things that do not have a lot of data modelling capabilities. This is the case for a class of gateways that are intended for mediating between different standards.
Do you think we are aligning, as an industry, on which language those gateways need to be able to speak? Do you think that we are moving towards a solution which can be reasonably expected to communicate in most instances, or will there just be multiple gateways and another layer that is going to mediate between these different gateways?
The market is huge. Parts of it are clearly headed towards a clear adoption of one standard, for example in autonomous vehicles for higher levels of autonomy. There are other markets much less established in the connectivity world because the challenges are either much easier or much more specific. For example, in the wireless activity space, there are definitely multiple different standards.
The connectivity framework that we chose is just another layer. The reason we chose that is that it is the most important factor for building a large industrial internet. The other parts certainly also need work, and it will be a longer time before there is really a consolidation of standards.
When do you expect that to be published?
I expected a few years because it takes a lot of work do a good job at analyzing the standards and coming out with useful information. It took us almost 3 years to do the connectivity framework; it took a lot of work to understand each technology to that level of detail. Meanwhile, the market is helping by choosing some winners and losers. The IoT market is so large that there is a niche for almost all the technologies out there, and the hard part is figuring out which one to use, since no one technology will be eliminated quickly.
I agree, the hard part for a lot of younger technology companies is also figuring out which market they should serve. I think a lot of people are still feel like they have a horizontal technology and it's a bit difficult to identify which which market do they compete well in, as opposed to having a kind of hypothetical solution.
I would actually recommend most technology companies to not focus on a single vertical because most use cases out there are not focused on a single vertical.
Take the analogy of animals: there are mammals that can fly like bats, mammals on land like you and I, and mammals underwater like whales. You could technically compete in all these different environments, but it is easier if you do a good job of understanding where you have real advantages being warm blooded and intelligent in the animal kingdom. You are not going to be able to compete in a space with octopuses, but you could definitely find a place where you could compete in all those niches. We at RTI made the clear decision that we are going to stick with our five questions, and that we will go after use cases that are not necessarily tied to every market. But in the markets where it is applicable, we are going to compete there. For instance, we have no use cases in manufacturing so far, but we have great use cases in transportation, medical, power systems, traffic control, and all sorts of defence systems, where we can compete quite well.
That’s a great point. At IoT ONE, we’ve just restructured our online database from industries and functional areas, to focusing on use cases to define what a solution does. That seems to be working a lot better.
I think you're right. It is much easier to understand the industry by its technical challenges than by its traditional verticals. It is very hard for people to accept that, because they are very used to going to conferences around medical or transportation, when they should really be going to conferences about highly reliable distributed systems.
We can probably learn a lot by talking to some other companies that are either maybe a little bit ahead of them from a different industry right. I think there's a lot of blind spots.
When it comes down to it, the whole IoT wave was really about technologies from many different industries impacting each other. All the confusion surrounding the IoT is because of that because there are technologies coming from different places where they’ve built a niche and are suddenly applicable to many other places becausecomputing power and connectivity to support that is now there.
It takes people who have been working in the power industry for 40 years, a long time to come to grips with the fact that the new technology disrupting their industry came from defense, or automotive, or something unrelated. That’s the real story causing all the disruption.
That and the fact that software is eating the world.
We have a growing number of companies that are approaching us and saying that they want to see what is happening in a sector that is completely out of their primary scope – things that have no direct relevance to their industry but are a few years ahead of them in terms of adoption particular technologies, especially in the consumer space.
There is a different consortia formed by ExxonMobil, a big oil and gas company, and Lockheed, a fence company, to build an architecture for refineries based on a standard called “Face for Avionics”, that is throwing the oil and has industry into serious potential disruption. A lot of people try to defend the old ways, but I think that is dangerous to do without really trying to understand new ways. The first reaction is “we already have this all figured out we did 20 years ago or 40 years ago”; the better reaction is “wow, how can I leverage these new technologies that are coming from other industries in my industry”. Very different.