Conversation with Dr. Richard Soley
Good day everyone, this is Ed Maguire, Insight’s Partner with Momenta Partners, and we’re here onsite at IoT Solutions World in sunny Barcelona. With me is the Grand Poohbah of the entire event, Richard Soley who is the Chairman and CEO of Object Management Group, but also the Executive Director of the Industrial Internet Consortium. I want to thank you for taking some time to speak with us.
It’s my pleasure.
I’d love to hear a bit about your background and what had brought you to the start of this event. There are many thousands of people here, we’d love to get a bit of a back-story on what brought you to create this amazing event in Barcelona.
Well, it couldn’t be in a better place first, but I’ll come back to that. We started the Industrial Internet Consortium as part of the Object Management Group, the OMG itself is 29 years old, I’ve been with the company 29 years today, by coincidence my 29th anniversary, I’ve been CEO for the last 21 of those years. And what we do is accelerate the adoption of various technologies, so OMG has always focused on standards, very quickly creating standards that have
But not everything is about standards, so when we looked at the IoT industry five years ago, we said it’s not ready for standards because nobody knows what’s necessary. What standards do we need to build? You need a sandbox, you need what we now called a testbed, but it’s basically a sandbox, a way to build stuff and see what you need.
- What are the best practices?
- How should you train people?
- How should you retrain people?
- How should you hire people?
- How do you avoid problems in your staffing?
- What kind of technology would make it easier,
- What kind of standards, and implementations of standards would it make it easier to build those testbeds?
So, we started four and a half years ago, this consortium, March of 2014, with the explicit idea that we would build testbeds in manufacturing and production, healthcare, automotive systems, other kinds of transportation systems, smart energy, smart grids, smart buildings, smart cities, and
We’ve since published four other testbed result papers on energy grids and so-forth, in our Journal of Innovation. But
Why this event? That seems to be a great question, considering there seems to be an IoT event every other day somewhere around the world, but this one is different. Not only is it in Barcelona, which is a wonderful city to be in, not only have we partnered with Fira at Barcelona which are fabulous partners, nobody is better, but it’s an event which is focused entirely on end-user case studies, even the keynotes about end-user, we require it, even the paid keynotes are required to talk about end-user case studies, and I think just as importantly, if you want to see, touch, feel, smell the testbeds, go to the floor, that’s where they are. Have you been there?
I have, and this is my second year but what really struck me about the booths was, how many live machine demos there are on the floor. I’ve been following what you define as IoT closely for about five years now from my Wall Street days. It’s a palpable change in how much we’re seeing real concrete demos in vendors booths now, not just a lot of slideware, vaporware.
I think that’s happening, that’s one part of it, the other part is a recognition that IoT is not in the usual sense a market. What IoT is, it’s something that impacts agriculture and healthcare, and transportation, and manufacturing and production, and energy grids, and smart cities, smart homes, and a million other things. Those markets are going to be irrevocably changed by IoT, so that’s what we’re showing off, that’s what you see in the booths, it’s certainly what you see in the test beds, and this you can touch and feel on the floor, how those markets are affected by the implementation of IoT; not just how the technology is used, but how the business models change.
The obvious one
People want the utility that serves their purpose and serves their interest, rather than owning part-assets.
Exactly, what Xentra calls an outcome.
That makes a lot of sense. I want to ask a bit about the process of putting standards together, and I’d love to get your perspective on what it’s like to develop a standard.
First, when you do surveys and ask people what’s impeding the adoption of IoT in your organization, they always say number one is interoperability, number two is standards, they’re wrong. Number one is talent, talent is far more important, and we’re developing talent in these programs as well of course, but these are people now that have built the IoT systems, and know how to build IoT systems. That said, it’s still valuable to have standards, and although IoT is not a standards organization, it is discovering what standards are necessary, and OMG very much is a standards organization, having developed about 1,000 standards over the last 29 years. There are many ways to develop standards, the most obvious of standards that were developed by a single vendor, backed by a single vendor, so Microsoft Windows, Apple IOS and so-forth, those are standards. You may not like how they work to you, but they are standards.
On the other
We do not require the step that the specification not actually exists, we require that it does
That doesn’t serve anybody’s uses.
No, it’s fun travel for people in the committee.
I don’t think it was intended as a
Well, having a boundary to the requirement, essentially it sounds like you have a distinct methodology to develop those standards, and then also distinct strict criteria that in short, they remain active and viable.
Exactly, and if they’re ever not implemented anymore, then the incentive goes away, it really does. We’ve done it three times, only three times but we’ve done
So, the Industrial Internet Consortium I know they’re doing a lot of work on reference architectures, and I think it would be helpful to provide a bit of context in terms of understanding how the consortium has come about, and how you think about dividing the
That’s a good question. First, how it started; five companies came together and realized that no-one’s going to win in this marketplace, either everybody’s going to win, or nobody’s going to win, but there’s not going to be a single winner, there isn’t going to be a Facebook, Apple, Google, Amazon. Why is that? Well, obviously the whole point is connecting systems that no-one ever intended to be connected across boundaries, across boundaries of applications, across boundaries of systems, across boundaries of countries. So, the whole point is connecting different systems from different companies, so you’re never going to see a winner, nobody’s going to win in the IoT space. I don’t even know what it means. So, what we do is, we start with the use case… well, backing up a little bit, we as I’ve said have a number of documents in our vocabulary reference architecture, connectivity framework, security framework, and
Those are about what you have to think about to build, for
So, an example of the used cases. The track and trace testbed, our very first testbed, and the first one that’s generating requirements for new standards, started with a very simple idea; if I knew where everything was in the factory, people, parts, work in progress, and tools, could I make the factory safer and more productive? I think that’s kind of silly because the answer is obviously yes, but it’s not so silly because you want to know how the answer is just yes. So, I think safety is kind of obvious, that is, you know where the robot arms are, you know where the people are, if the robot arms are about to swing around and kill somebody, you can stop the robot. It’s a real problem in factories, and what they normally do is they put robots in cages so that people don’t accidentally go near them. Did you know that 90 percent of people hurt by robots in
Really, is that the complacency?
Exactly yes, they know better. So, that’s a real issue. From
So, just to give you an idea of other testbeds that look completely different, we have a testbed also run by Dell in Southern Ireland, County Cork. Again,
How do you look at capturing best practices? Because it would seem, just in the two cases that you’ve described in the factory and ambulances, there would be a wealth of insights that come from being able to collect all this data for the first time. Do you have a structured process to codify the best practices?
I wish we did.
Not that there necessarily is an easy way to take a cookie kind of approach to such different situations but, how do you think about the scenarios that evolve. I’m interested in understanding the iterative process, and how you arrive at some of the insight that
To date, we don’t have a very structured approach to reporting best practices, I wish we did, in
You’ve brought up a really important point, which is the constraints of talent that exist in the industry. In our
Flexibility versus robustness.
That’s right. How do you envision in an ideal way how the output, what you’ve learned, how these testbeds will help develop
The initial way is trivial, that is the people that built the testbeds as part of the testbed development project, have that expertise. What you’re really asking is the harder question which is, once we’ve collected best practices, how are they taught? And that’s why I really want to work with
It’s a question that keeps coming up, and I think in many respects it really reflects how nascent the discipline around Connected Industry, I’ll use the term industrial IoT, really is.
You can use whatever term you like, the Germans use industry
Well, I asked the head of cyber-physical systems at NIST, ‘What in the world was a cyber-physical system?’ He
It’s not a very good marketing term either.
No, well it’s not. We use industrial IoT for a very simple reason, there was this revolution from 1850 to 1950, it was the Industrial Revolution, productivity increased by a factor of four during that time by the way, and that productivity increase created a large increase in consumer demand, which created jobs, thank goodness, because jobs were lost to industrialization. It happened again from 1960 to 2000, that was the Internet revolution which moved human connectivity to machines, and during that time
It’s really fascinating just walking the floor here and seeing how many industries are rethinking just fundamental processes. What’s surprising to me is in a sense how quickly a paradigm vision can take hold, for instance, the whole concept for transportation services, and the reality of how close some of that technology is. What are some of the significant shifts in looking at the way companies and end-users – I guess the stakeholders that you work with, have looked at the market over the last five years or so? The reason I choose that time period is, that was when we got the initial excitement about IoT in 2012-2013, vendors were started publicizing big forecasts of economic value-add, and connected devices. But I would love to get your perspective on the evolution of the practical perspective.
I think despite what the Wall Street Journal says, the average person doesn’t care about IoT per se, it just wants safer minds, better citizen services, more productive factories, more productive farms, and
I bought a memory card in 1976, I’ll never forget, a triple-high VME card, 64 kilobytes memory, $16,000. 64 kilobytes of memory. Now in my bag sitting over there, has 2 gigabytes of memory card, just in case I need it, and I think it cost me $9. Process and power
I built a voice recognition system in 1981 when I was at MIT, it used the mainframe, $10 million
In other words, Arthur C Clarke said, ‘Any sufficiently advanced technology is indistinguishable from magic’. I think he has a point, I think people don’t care about the technology per se, they just want the outcome.
Technology vendors though tend to get a narrative, bright shiny objects, early-on…
And you want them to! That’s critically important, otherwise, that stuff is never going to get developed. So, I’ve been on both sides of that divide, and let me tell you, if nobody got excited about bright shiny objects, nobody would ever build bright shiny objects, so, it’s important. I think vendors do a great job of understanding what people want, and how to implement it with the technology that’s available.
With the makeup of this conference, I don’t have much perspective because this is only the second show that I’ve been, but I’m certainly seeing a theme emerge where there is less of a focus on the sensors, and some lower-level gateways, and a lot more focus at least in the rhetoric that I’m seeing, and please correct me if you think differently, I’ve seen a lot of focus on business outcomes.
Exactly, here downstairs is a room that’s just talking about mobility, there’s a room that’s just talking about healthcare, there’s a room that’s talking about… we don’t have agriculture, but you get the point. But those rooms are separated, and the program is designed by vertical, not by technology, we don’t have a room where we talk about sensors in a room, or we talk about actuators in a room, or we talk about processing power. Frankly, that’s not what people want, they want outcomes, so give them outcomes. Equally, we make sure all the speakers are talking about how it implements some particularly important used case, not what technology is cool, not the cool of new reference architecture.
That’s critical, the sessions that I’ve been, it’s a really high-quality group of speakers, and I think Jessie DeMesa has been involved from Momenta in putting together some panels, and I have to say it’s a different tone of a tech-focused conference, at least in the last couple of years than what I’ve been seeing over the first round that I used to go to.
And, more and more every year. Last year we required that the non-plenary events focus on end-users, now we required that all the speakers focus on end-users. I think that really works, it makes the diamond sponsors unhappy sometimes, but it makes the attendees to the event happy, that’s what I care about. I think the fact that we center the exhibit floor on testbeds sends a message, that the most important thing is real systems to solve real problems. It’s kind of hokey to say it, but the fact that there’s a fire truck down there, says, I don’t want to care about sensors, I want to care about fire trucks.
There’s an interesting presentation on shoes, that Ecco the shoemaker has this ability to analyze someone’s gait to create a better insole.
Yes, exactly. You look at some of the keynotes, I didn’t get to go to many of them, because I’ve been doing things like this, but for example, Jonathan Ballon for example at Intel, he had up on stage, one after another, end-user case studies, and he was talking about what the case studies were, and how the technology implemented those case studies. That’s what you want to know.
Look, every engineer, and I is one! … starts with the problem of, ‘What do I do first’, and a lot of what you hear answered downstairs is what you do first… and it’s not obvious. And you’re going to learn that from people who have already done
Are there any industries that stick out to you, at least in your experience, that have been particularly successful in being able to demonstrate the potential of…?
Everybody asks which industry is going to adopt IoT first. I actually know,
So, I really think it’s going to be manufacturing, but I will tell you what the other thing it is, the job of a taxi driver in 20-years is going to need to find another job.
That was a key point and a key concern that again, you highlighted earlier, we’ve had a couple of cycles where there was
I thought you were going to say overblown. I think they’re not overblown, I think their worry is perfectly reasonable. During the industrial
When there was only one Webmaster in the world, his name was Tim Berners-Lee, nobody could have guessed that the job down in five years would be Webmaster, and there are now tens of millions of Webmasters worldwide. Those jobs were created out of the blue, but more
I was a moderator on a panel, and the discussion was how automation and artificial intelligence are refactoring the existing jobs within factories, it seems there’s still quite a bit of cultural resistance…
Yes, but it’s like holding back the tide. There’s a great book that’s sort of on this topic, it’s called ‘The Box’, and it’s a history of containerization, it’s about how containers were funded, standardized, how ships, trucks, and airplanes and so-forth started using containers. How we ended up with the standardized 40 ft
Sounds very typical.
With the net result that people shipped into Oakland which was not unionized, even though they still had to ship back across the country from Oakland to New York. The net long-term effect was the creation of the Elizabeth, New Jersey, completely automated shipping systems, in New Jersey
That’s great that you
That is amazing, and that’s a great way to circle back again to the value of standards. The Romans standardized the width of the carts…
Oh, it’s such a good story, the standard gauge for trains in the United States is 4ft 8½ inches, that is a bizarre gauge. Why is it 4ft 8½ inches? And by the way, if you work it out in metric it’s not a good metric either, that’s because that was the standard gauge after the 2ft attempts in the UK as well. Why was it 4ft 8½ inches in the UK? Well, because the first rails were laid on the flattest roads that existed in the UK, which had been built by the Romans. Why did the Romans choose that width? (that’s twice 4ft 8½ inches of course, because there’s travel in two directions) Well, that was the width of a Roman war chariot. Why were the Roman war chariots 4ft 8½ inches wide? Because that’s the width of two Roman war horses, yoked together. So, now you know.
But now move forward, the Space Shuttle had two Thiokol boosters, which were a little bit smaller than intended, that’s one of the reasons they had to have that huge booster in the middle. How did the Thiokol move those boosters to NASA’s launch sites in California, in Florida? By rail, and that meant that the boosters had to fit under all the bridges, and across all the bridges, and so forth on the rail systems of the United States, so they could not be more than 4ft 8 ½ inches around.
That is an amazing story.
Which tells you that one of the most modern transportation systems in the world, the United States Space Shuttle was designed partially around the width of two Roman war horses butts.
It tells you that the standards, these rails of course which give rides to businesses and commerce, and communication, right now the standards that are emerging today will be the railroads of commerce and innovation for centuries to come.
Absolutely, and technology systems are far less important than other considerations sometimes, for example, the two newest aviation plants in the United States, one in Alabama, one in Charleston, the reason they’re in those locations is
It’s amazing. We’re coming up to the end of our allotted time Richard, and I want to thank you so much for your insights, it’s been fascinating. I’ve
Mostly the team which makes me look good, they do all the work! And I thank you for doing panels, and participating in panels as
I’m very passionate about it, and so is our team. This has been Ed Maguire, Insights Partner at Momenta, and we’ve been speaking with Richard Soley, who is the CEO of The Object Management Group, and Executive Chairman of the Industrial Internet Consortium, but also the driving force behind the IoT Solutions World Congress, here in Barcelona. Thank you again for your time.
Thanks for coming.