Podcast #76: Industrial Insight Transformations

Good day everyone, and welcome to another episode of the Momenta Podcast. And today this is Ed Maguire, handing over the role of Insights Partner to Leif Eriksen, who has a long career as an analyst, as a writer, as a thinker, and it’s a pleasure to have you onboard Leif.

Thank you Ed, it’s my pleasure. 

So, I’d love to start with a bit of background, and have you share a bit of the experiences in your career, could you tell us a bit about where you started, and what had gotten you interested in digital technologies.

Yeah, absolutely, love to do so. So, I started over 30-years ago as a young engineer, deploying and configuring instrumentation, automation, and advance process control systems in the petrochemical industry. We sometimes think with all the hype around the industrial Internet of Things, and AI etc. that the use of digital technology in the industrial sphere is relatively new, and nothing could be further from the truth. We were doing some very interesting things back then, obviously there was also some significant constraints, one of which was we had a lot of disconnected systems.

So, to give you an example of that, we would collect all our data from the sensors that we had, they were all wired up to a control system and then that data was imported to another computer, in our case it was a Deck PDP 11 minicomputer. This was the dawn of PCs and applications on PCs, and so we got to do some interesting things, but there were also again some significant challenges, one of which was, okay how do you get the data from that minicomputer to the PC, when you didn’t really have a network to do that? So what I had to do back then was, throw out reams and reams of paper on a dot matrix printer from the PDP 11 of data, and then manually re-enter that data into the PC, into what was at the time a novel program called Lotus 123, that was the dawn if you will of the ubiquitous spreadsheet era, and then look to do something with it at that point.

This is when I learned my first and maybe hardest lesson on data backup, I spent a whole morning literally manually typing and entering data into the PC, the PC crashed and I hadn’t backed up on what was at the time the 5 ¼ floppy disk that was from the 20 megabyte hard-drive, to the 5 ¼ floppy disk, and so I had to start over again. But what got me excited, what was really exciting about all that was, once you got that data into the spreadsheet you could do interesting things with that; you could look at trends, you could compare process data with advance, you could even do some fairly sophisticated statistical analysis. So, the kind of insights we got into what was going in the plant, and some of the issues we had, or how we could optimize or improve the performance of the plant was very interesting, and very exciting.

So, that’s what got me hooked, and to this day that’s what I’m passionate about is how do we use digital technology to truly optimize our industrial operations.

So, you spent a lot of time with your sleeves rolled-up, really understanding how the technology worked from the ground–up, and you leveraged that into analysis. I would love to get your perspective on the transition in your own experience, as you went from really being operationally in the weeds, to taking a much I would say broader view of the industry and technology. Tell us about that transition and some of your experience there.

Absolutely. So, 15 years into my career as an engineer in the oil and gas industry, in which I had dragged my young family around the globe; we lived in Canada, Texas, the US Virgin Islands, and eventually Indonesia, and an opportunity came up in 1998 to settle down if you will and do something different, and joined a small analyst firm called AMR Research, prior to that Advanced Manufacturing Research. What got me excited about that is, in that role I could look at the broad brush, the broad scope if you will of the use of technology, not just in oil and gas but across a lot of different industries and get a better understanding of what was state of the art at the time, and what was coming, and so that was very exciting.

So, making that transition was both challenging but interesting, and so it was really the beginning if you will of the next 20+ years where I’ve really dedicated my life to helping companies make sense of technology trends, and how they impact their business, and what is it that really matters. There’s a lot of hype out there in terms of the various different technologies, today of course a lot of its around AI, machine learning, blockchain, but really how relevant is that to my industry? That’s what business leaders care about, and that’s what really I’ve dedicated my career to.

What are some of the most impactful shifts in technology, or even in thinking, over the last couple of decades, as you look at how… you started working in truly industrial technology environments, but as you’ve seen this parallel evolution of information technology, are there some key trends or developments over the years that you can point to, that have created real inflection points for changes in thinking, or really innovative applications of technology?

Absolutely. In some ways a lot hasn’t changed, if you think about what engineers are doing today in plants around the world, well a lot of them are taking data out of various different systems that come in from sensors etc. and analyzing it, and it’s not unusual for them to still use a spreadsheet. But they also have more sophisticated tools available to them, but really what I see as a major change, and I started to think about this back around 2001 and 2002 when I was still at AMR Research, that was that with the internet there were all sorts of possibilities that were opened up in terms of what you could do with the data, how you could share the data, how you could collaborate on the data, how you could analyze the data, and ultimately how you can make decisions based on that data.

Ultimately the goal for any company is to make better decisions about how they operate and maintain their assets or their equipment, any industrial company in any case. Everybody wants to get to this Nirvana if you will of the perfectly predictable operation, where they have enough data and they’ve got sophisticated enough analysis that they can say that this is what’s happening now, with a high degree of certainty, and this is what’s going to happen tomorrow, a week later, or two weeks later etc. So, we’re getting closer.

But I started to think about it in 2001-2002, and the basic tools were being put in place back then. But what I didn’t realize, and maybe at heart it’s because I’m a bit impatient with these things, is that it was going to take longer than I thought. Part of that of course was, if we look back there were two recessions in that area, we had the dot com bust and recession that followed that, and then of course the great financial crisis of 2009. There’s also the issue of some of the changes that this technology will ultimately create in the business world, in the industrial world, are quite disruptive. So, people inevitably dig their heels in when something like that comes at them. I guess the final thing is that probably the cost benefit equation back then wasn’t quite there, and that’s certainly changed over the last 15+ years.

Yes, I think you articulated a couple of key points here, one is that there’s these principles about taking data from systems to optimize operations, reduce risk, and grow sales, are really very much the same as they were 30 years ago I guess, in kind of the dawn of the data warehousing age, but the technology of course has improved, and the tools have evolved and are more powerful. But as you look at where we are today, what are some of the things that get you really excited about the state of the market today, and where we might be going?

That’s a great question, and really it comes down to the maturation of the cloud, and cloud tools available in the cloud. A lot of that credit goes to the big consumer cloud companies, whether that’s Amazon, Facebook, Google, or to a certain extent Microsoft, they sort of straddle both the enterprise and the consumer world, but they’re massive investments and their Azure infrastructure has also been a major factor. If I think about it, if I go way back when to when I first started, it wasn’t that unusual that you had more sophisticated technology at the office than you had at home. Again, it was the dawn of the PC, we started to buy PCs for the house, but they were relatively expensive, and they weren’t that sophisticated, so the odds are you had more sophisticated technology at work.

Well, that has completely been flipped on its head, and largely as a result of the investments made by these large technology giants. If you think about what we can do today with just a mobile device, and iPhone or similar, in terms of controlling our home remotely even from remotely purchasing basically anything, anytime, anywhere; this kind of capability in some ways is not quite there in a lot of enterprises. A lot of that has to do with the fact that companies still store their data in local functional silos, it might even be as local as a single plant, but even if it’s sort of captured up at an enterprise level they’re in functional silos and the systems are hopelessly incompatible, so what gets me exited is the ability to take that data and put it in the cloud, and do all sorts of things that you can’t do. But most importantly, the ability to share that data in real time with a broader ecosystem, with your partners, across your own organization.

When I was at Gartner recently, I spent six years at Gartner from 2010 to 2016, whenever I would have a call with one of our customers, our end-user customers, I would joke with them that, ‘Information moves faster outside of your organization than inside’. So, you could have something happening in one of your plants on the other side of the world, and the way you hear about it is over the internet, not from your own people or your own systems. So, the value of putting that in the cloud to both internally share it… I’ll relate another anecdote, I spoke with a large global chemical company, and they started probably as long ago as six or seven years to use one of these cloud-based services, because they had 100 plants around the world, but it was the best and most efficient way for them to share best practices and information. So, that’s going to have a huge impact on how decisions are made, the quality of decisions, and the performance of these facilities.

Could you talk about what are some of the characteristics of companies from an organisational perspective, that are successfully adopting a cloud model. Because I think when you have alluded or what you’ve referenced in this, again where you move data out of silos into a cloud model, also has a lot of governance dimensions and changes to process, and the ability to relinquish control in many respects right, of people within an organization that have done things a certain way. What are some of the characteristics of a legacy company that’s been able to make a successful transition? And on the flip side what are some of the obstacles that you see to being able to completely embrace this newer paradigm?      

Well, you hit the nail on the head when you said, ‘loss of control’, because when you start to talk about things in terms of putting data in the cloud, ensuring that data more widely, there is a bit of loss of control, because information is control and power in an organization, and a lot of the structure of the modern organization, the hierarchical structure, the functional silos etc. are built around information silos and information control. But I want to be careful with the word ‘control’ as well, because in the industrial world unlike a lot of other industries, control also means control of equipment, and that’s something that people aren’t prepared to, and probably shouldn’t be considering depending on the-the complexity and the risks associated but putting that in the cloud. So, most companies are not going to go that direction, simply because… or at least not into the public cloud, and there’s variations on that, but simply because if you operate a plant or a facility where there can be a major safety incident, or there could be a major environmental incident, there’s a lot of liability, there’s a lot of obviously human risk and human consequences in doing that.

So, there’s control, and there’s control; there’s control over the information which best in class companies are willing to give up. Some of the companies that are providing services in this area… sorry, you’re talking about the liberation of data, you interviewed Francois Laborie of Cognite not that long ago, and that’s their moto, ‘Liberate your data, put it in the cloud’, and so being able to give up that control and to recognize that when you share that information more broadly, you’re likely to get a better result in terms of decisions, and decisions will be made much more quickly.

One example I’ll give you, we’re all familiar with the Deepwater Horizon catastrophe, and that was a catastrophe on so many different levels in terms of the loss of life, and the environmental consequences, obviously there was a huge financial consequence to the companies involved. There’s a tendency to view it as an act of negligence, and in reality, this is an industry that probably spends as much money and effort on safety, or more money on effort and safety than any other industry, maybe with the exception of the space industry. The problem is of course, these are very complex operations, and there’s a lot of things that happen, and they still rely on local expertise.

So, let’s imagine a scenario where all of the data, all the decisions being made related to the drilling of that well were shared widely across BP, across their partners etc., and everybody was able to look at it. Not only that, at some point being able to subject that data to some very sophisticated machine learning, where it might say, ‘Well if this decision was made, and yet the pressure was this in the whole, or the pressure really doesn’t seem right, we need to go in a different direction’. So, the ability and the willingness to share that data I think is really what distinguishes best in class companies, and there’s very few of them today; it’s just such a hard thing to do to give up that control, and in some cases there are illegal and regulatory constraints to doing so.

I think you are touching again on another really important theme in digital transformation, is where the decisions lie, and what are some of the key decisions that need to be made to really affect what we will call a digital transformation, which is this move to this broader model of agility and sharing. So, where do you see some of the key decisions being made for industrial companies, is it a Chief Digital Officer?

Well where there is one, not everybody has one today so that’s kind of an emerging role if you will, but where there is one it may be there, it might be the Vice President of refining, the Vice President of manufacturing, ultimately though it has to  come from the CEO, the direction and the vision has to come from the top, simply because the transformative impacts are so huge. If you think about you start to put data in the cloud, and you start to expose it to a broader audience, then that completely changes how decisions are made, and who makes them. That then completely changes what your organisational structure looks like, what sort of resources you need on the ground, if that data is at the very least being made to… if everybody at HQ for example has complete visibility to all the data, and all those facilities around the world, and they’ve got some capability; whether its visualization or analytical tools to put some context around them, help them make decisions, what are the implications for the engineers and others, supervisors, on the ground who have typically made those decisions?           

There’s probably a role for some of that expertise at HQ now, but less of a role at a local level, and so again if you think about the traditional organisational structure, it’s all built around, ‘Okay, let’s flow information and advice up the organization, multiple levels, and we all know how long that sometimes takes, and how often context is lost along the way, and then the decision will come back down from the top, what we should do. So, when you take that model and you circle it, let’s blow that up, let’s just make the data available instantaneously to everybody, what does that mean for the organization? Most organizations are not well-positioned to deal with that transition, they’re not ready to deal with that transition, and that’s why it has to come from the top, because the threats across the organization to different functional groups and different leadership structures is fundamental, and very disruptive.

Well when you start to automate processes, one of the challenges for anybody in a decision-making role is the risk that with the adoption of automated systems, predicted systems, that takes your power away from you. When you have to exceed the control of a decision to a machine or an algorithm, it’s obvious that there would be some sort of resistance to that. I’d love to get your thoughts on how organizations should anticipate meeting with that resistance, how to articulate a successful vision, and ultimately how do you get past resistance and really get buy-in across the organization, when you realize there are going to be changes in roles, and just the way that jobs themselves may become refactored.

Ed, that is the biggest challenge, and probably the reason that in the industrial sector specifically, the change hasn’t come about at the same pace as it has in other sectors, such as retail and financial services. Retail in particular as we know, Amazon and some others but primarily Amazon is a huge disruptive force, one that incumbents have had to react to if you will and fight back. So, they didn’t have any choice, you don’t have that same dynamic in the industrial sectors because there’s a lot of some cost if you will into plants and equipment, and these plants and equipment can be very complex and expensive to build, to operate, and to maintain. So, it’s not like it’s simple enough for a newcomer to come in and say, ‘I’m going to build the next great oil and gas company’, or, ‘The next great automotive company’. Putting Tesla aside, they still have to build the plants so to speak, so that’s a different kind of disruption if you will, maybe less digital than the energy base.

The bottom line is, for the industrial world change has been slower because it’s been easier to sustain the status quote, and in some cases, it’s been also driven by the regulatory constraints and requirements. So, there’s been lots of different ways at which customers have been able to maintain the status quo, dig in their heels if you will, but it’s coming. We see a number of different initiatives out there where visionaries from industry are out there raising funds, pools of money to create the next great oil and gas exploration company, one that’s based strictly on data and analytics, and has a fraction of the number of people involved in decision-making process. Same thing on the manufacturing side. So, it’s going to come, it’s just going to take a while to come.

The other thing I would say is, when regulatory authorities start to recognize that industrial companies can invest in technology to reduce the number of incidents that might occur, and improve the performance and predictability operation, if the companies themselves don’t embrace it then it might get mandated. So, we don’t have the industrial accidents where people are maimed or killed, we don’t have environmental catastrophize like the Deepwater Horizon.

Yes, it sounds thought that there’s enough of an organisational or cultural inertia dynamic, that you need some kind of push or some catalyst to really drive an inflection of real disruption in the industrial world. I’d love to get your view on what could be these catalysts, what could tip the scales or be the proverbial strong camel’s back that really drives an accelerated adoption, and transformation in the industrial world.

Yes, I know, it comes down to Ed the data going into the cloud and what that allows companies to do. So, the companies that embrace that, again to use the cognate term ‘liberate their data’ and put it in the cloud, and start to collaborate around not just internally but with a number of their business partners, their suppliers, etc. So, imagine if you think about most manufacture industrial operations are an amalgam of a variety of different pieces of complex equipment that are engineered to work together. Those pieces of equipment are not manufactured by the operator themselves, but by a third party, well who do we think has the most expertise? In fact, typically if there’s really a problem with a piece of complex equipment, the operator gets on the phone and calls the manufacturer and says, ‘Can you send one of your guys down as soon as possible?’ Well, imagine if that person could access that data, was actually monitoring that data in real time for their customer, and not waiting for that call but pre-emptively sending them a message and saying, ‘You’re going to have a problem with this piece of equipment in the next week or two, if you don’t do this or that’.

So, that completely changes the paradigm in the business model, and now you start to say, ‘Okay, what does the operator or owner if you will of the assets of the equipment do going forward?’ well, they’ll outsource more of those kinds of functions, and they’ll outsource it to people that truly have the most expertise in terms of both monitoring and maintaining the equipment. So, that model again; once that data gets out there, that’s what’s going to create… and so there’s two scenarios here and it’s going to vary by subsector industry; one scenario is visionary companies with visionary leadership will start to do that and make those radical shifts or changes in their own organization, and how they run their business, and they will force others to follow suit, because they will be again radically more efficient, radically more reliable, more predictable, and that again is the ultimate goal.

The other scenario is you’ll have some outsiders come in, an again some industries that’s the way it’s going to happen, if outsiders can raise the capital to build the next generation of company in that particular sub-market. Again, oil and gas exploration comes to mind because they’re familiar with a number of initiatives out there where they’re looking to build again from the ground-up, something that’s radically different than what was done before.

One of the big challenges of adoption of digital technologies across industry has been, and I still see us popping up in surveys as a top obstacle to adoption, and would love to get your sense; we’re about a decade out from the first wave of operational technology attacks, I think Stuxnet was 2009 or 2010; what from your perspective are some of the key developments relative to security and operational technology, any other developments that you think are constructive for some of these later adopters to feel more comfortable about jumping head-first into digital business?

Very important issue and top of mind for a lot of executives in the industrial space, and what the Stuxnet incident taught us is, in the OT world particularly, you have to combine physical security with cyber securities. So, there’s still the option in if you don’t have good physical security of lock-in with the USB stick in your pocket and creating the disruption that way or the damage that way. So, that’s certainly one of the things that distinguishes it.

The other thing that is sometimes lost in the noise around cyber security is, in the IT world the primary goal is protection of information, protection of information about people, about things, privacy, that type of thing. In the OT world the primary thing is protection of equipment, and the control and operation of equipment, so another way to look at it is the reliability side of the machine. When you think about security in the OT world you have to put that first and foremost; it’s less about what information you pump out into the cloud and share, and more about how likely is it that someone can come back in through that tunnel or another tunnel, into your control systems, and change something which causes the plant to shut down a major operational disaster etc.

So, it’s a very interesting area, in fact as you know the next podcast we have coming up is an interview of Heather Ingle, the managing partner at Cyber Strategy Partners, and we’ll hear some of her perspective on this, but it’s still an evolving area, top of mind for all executives in the industrial space. And going back to the point I made earlier, that’s why for most industrial operations its not appropriate to put the actual control of equipment in the cloud. Data, yes, control, no.

At the recent IoT Solutions World in Barcelona there was an interesting trend, a lot of the traditional industrial players like Schneider Electric and ABB were not there, but where there’s a growing presence from traditional IT vendors like Microsoft, Google, and others, what do you think is going on? Is there a shift in the industry here of power, or maybe just the state of tradeshows?

It could be either of those, that’s a possibility as well, it’s hard to read so much in there because there’s too much intuit because there’s so many different demands on resources from a marketing perspective of these organizations. That said, this is a significant event, and the fact that it’s the first year that they really weren’t there at all, I think it is significant; and maybe just as significant is that the big IT players were there, and certainly they’re reaching further into the enterprise if you will, and reaching further into the OT world with some of the technologies they’re offering. Of course, Microsoft has always had a substantial presence in the OT world, a lot of the systems that the big OT vendors sell is based at its heart in Microsoft technologies and OS’s or variants of them.

Yes, I think it is, it is significant and it also may reflect the fact that one of the things that I’ve been noticing for a number of years now, let’s say for the next five or six year, is that a lot of the technology spin in large enterprise and particularly in the industrial world, is shifting from the IT side of the house, to the OT side of the house. Why is that happening? Predominantly because IT has already spent the money if you will, the earpieces are in place, financial systems are in place, inventory etc. what’s left to do except possibly move those systems into the cloud and make it more efficient to operate the system, in other words to cut the cost of running these IT systems.

Ultimately where the investment is going to go in these organizations is into the OT side of the house, because a lot of organizations still have fragmented systems, different systems in different plants, and as I talked about earlier, hopelessly incompatible; you’ve got a maintenance management system that’s not communicating with your data historian, someone’s got to bring all that data together. So, those are the game changers that’s going to happen in the cloud, and again that will be the transformative trigger if you will for the industrial world.

Do you see a shift in traditional operational technologies, for instance plant floor data-management, do you anticipate new types of technologies, or new approaches evolving that will be information technology, but specifically tailored to heavy industry?

Yes, so the primary shift there is, if you think about it, go back to the beginning of our discussion when I talked about starting my career, and that I was able to take all this operational data, put it in a spreadsheet and analyze it, that was a very local process. Interestingly enough, that’s still true in a lot of different organizations around the world, they’re still done locally by a local engineer making that decision. The other part of it that hasn’t changed dramatically is that they’re still predominantly looking at processed data, and they’re not overlaying that on top of maintenance history, or maybe some new types of data, audio data, or visual video data.

Audio data for example of a noise that a machine is making, and of course in anything where weather is a factor, bringing weather in. So, the ability to take all of that data, again put it in the cloud where it’s much more efficient to manage, share, collaborate around, analyze etc., and put it together in a way that delivers context, and then maybe ultimately providing overlaying some machine learning, or artificial intelligence. That’s going to change everything, it doesn’t mean you’re not going to still be collecting the data the way you do today, you may or may not depending on whether there’s a need for more data than you already have. But it means that how you look at the data, how you analyze it, and where you do that is going to change dramatically, and that’s going to have an impact on the OT vendors that we were talking about earlier, is are they going to be providing that capability? Or, is it going to be one of the big IT vendors such as Microsoft or Amazon with AWS, or is it going to be some startups?

There’s actually a very-very large goldrush going on right now where billions of dollars are being thrown at startups and invested by incumbents chasing this opportunity if you will.

So, we look forward and there’s a lot of moving crosscurrents in IT, operational technology, and industry of course, and as you look forward a few years, what gets you excited, and what are some things that keep you up at night?

Yeah, what gets me excited, and maybe it’s the same thing that keeps me up at night, is the possibility that we can achieve this Holy Grail if you will of the perfectly visible, perfectly predictable operation in the industrial world.

We’re not that far away from that, the technology in many ways is there. The biggest challenge quite frankly is not applying some sort of sophisticated analytics to the data; it’s collecting the data and contextualizing the data, and then sharing the data. And for most organizations today, that’s impossible if bot ridiculously expensive to do internally, so the only viable option and the inevitable option is to move it into the cloud, and that’s what gets me excited because the capabilities that exist in the cloud today, the ability to do that, the costs – barriers to doing that have dropped significantly that we can envision this world where unplanned incidents don’t happen, people don’t get hurt in industrial operations, partly because they’re not even out there anymore, but also because we know what’s going to happen and when it’s going to happen.

So, we’re closer than ever to that Nirvana, and the sooner we get there the better. One of the most famous organisational consultants of the modern era, Warren Bennis said back in 1991 that the factory of the future will be reduced to two living things, a man and a dog; the man will exist to feed the dog, and the dog will be there to make sure the man doesn’t talk to the equipment! What he was basically saying back then is that the autonomous factory, the autonomous industrial operation will come. He visualized that back in ’91 and we’re I think finally on the cusp of that.

That is really exciting, and we see the changes in every industry, in companies of all sizes as well. It’s been a fascinating conversation and I just wanted to wrap up with some thoughts as you assume the role of Insights Partner at Momenta; what are some of the key themes and areas of focus that you want to be highlighting, and areas of particular interest to you as well?

I couldn’t be more excited about joining Momenta because it’s going to give me the opportunity, and us the opportunity to be at the forefront of this transformation we’re just talking about. This is a scary challenging difficult time for a lot of leaders in the industrial world, and so navigating the journey from where they are today, to this brave new world is going to take a lot of vision, but it’s going to also take a lot of guidance. And so being able to rely on an organization like Momenta, to provide the kind of insights that you need to make the right decisions, both in terms of the overall strategy, but also in terms of the particular technologies you should invest in, in any given point in time, how mature they are.

As I mentioned, there’s this massive goldrush going on, and there’s no shortage of companies or vendors if you will, start-ups and established ones, banging on the doors of the Chief Digital Officers and other leaders in the industrial space, to sell them the next great thing. Well, is it really, do those companies really understand the Chief Digital Officer’s industry and even its own company?  Industries are very different even within the context of the industrial space; what they care about, what’s important to them is different. An asset-intensive organization like the energy industry, or mining or whatever, it’s about the reliability of equipment, the ability to run that equipment 7 x 24, 365 days of the year.

For a food and beverage manufacturer it’s all about quality and compliance, so you want to make sure that the food… and it’s also about in this day and age appealing to increasingly localized tastes and requirements, and so being able to do that requires a different set of tools and technology, requires a different strategy, so we’re going to be out there not just talking about digital transformation and digital industry in the industrial world in general, but also talking about what it means for individual companies and sub-sectors. So, that’s what gets me excited, is that we’ll be leading the way in that regard and look forward to talking to a lot of companies, a lot of Chief Digital Officers about their challenges and how we can help them.

I’m excited to see your work and hear more from you, and I want to extend a warm welcome from the whole organization to have you join.

So again, Leif Eriksen welcome as the new insights partner at Momenta Partners, and again, this has been Ed Maguire with my final episode as hosting the Moment Edge Podcast, and it’s been a wonderful run and I’ve had some fantastic conversations. I couldn’t be more pleased to have you taking the reins and sharing your insights and help them grow the future. Thank you very much.

Thank you Ed.

[End]