Saurabh Gupta — President, HFS Research[00:21]
Welcome to today’s episode of HFS Unfiltered. And today, I’m excited to talk about the convergence of the physical and the digital world, with none other than Piyush, who heads STX and AI at LTTS. And, by the way, LTTS just came out as a market leader in our most recent IoT Horizon. So I’m delighted to have you here, Piyush.
Piyush Jain — Head of STX and AI, LTTS[00:47]
Thanks, thanks for having me. It’s my pleasure to join on this podcast with you.
Saurabh Gupta — President, HFS Research[00:52]
So, Piyush, tell me — you’ve been in this field for so long. What’s been the most exciting IoT use case you’ve worked on, and why?
Piyush Jain — Head of STX and AI, LTTS[00:52]
We have been working for many years now, because IoT is not new. It was something that possibly started somewhere in the 2010s and 2012s, where Cisco came out with this concept of the Internet of Everything. And I think it was Industry 4.0, or a predecessor to it, that kind of took it off. So we have been actively engaged on this for many years now, and it has matured. Case studies, if I have to put it that way, which have been quite challenging but very interesting work that we did. One was for a plant engineering customer where they wanted their critical assets in the plant to be instrumented, monitored, and managed, and they were looking for an IoT kind of solution where a lot of data was getting generated by these critical assets. It was a complex plant layout, so they wanted to see how to harness all the data that’s coming in onto a platform, and the challenging part was not that. The challenging part was how do you model this different equipment. What are the critical parameters that you’re going to monitor, how are you going to model the parameters and the data being acquired, and then model what you deduce from that data, right? So it needed not just the technical expertise of how you lay out that entire IoT architecture, but also a lot of domain expertise to understand the parameters that are coming in, how you put them into the model, how you construct the model to decipher the patterns, and so on. So it was a pretty long exercise for us to even prove to the customer that this is going to be useful. Because they had their doubts. They wanted to experiment with this entire concept of IoT and how it can help them monitor the equipment, bring in preventive maintenance and predictive maintenance kind of functionality to their plant. So it took us quite a long time to really get this structure in place, get the parameters out, iterate through the models to see how those parameters make sense, and once we could prove to them the kind of data that was coming in, what it was telling, and how we could actually save their downtime. We did it for one plant to start with, right? And once they started seeing that outcome is when they started believing in it, and then they allowed us to roll it out into a number of plants thereafter. But that entire exercise was a learning curve for us as well. So for me and my team, it was also great learning to understand the industrial side of things — how the equipment is actually working in the field, the kind of data it is generating, what it implies, what downtime for a plant truly means in terms of the business impact it creates, and so on. So I would say that was a very challenging exercise. It was a long project for us to get done, but it was very satisfying in the end. We could actually give the business benefit that the customer was looking for. The second use case was for an industrial customer who is into the lighting business. And they were developing smart lights. This was already a smart product, but their whole goal was, as this lighting gets into a city installation, they wanted to see how to smartly monitor it to save carbon footprint. So the goal was to really do the energy saving and prove it to the municipal corporation of the city that by smartly managing it, you really save carbon footprint. So the challenge for us was to make sure that the model we were constructing on the platform was such that when there is no one there, the streets are not completely dark, because that’s not what the city would want, but they’re dimmed out, and when there is someone there, it brightens up again and dims out. And how do you configure how much dimming and brightening you need to do? So all of that again was very interesting. It was a little bit of a complex layout, but then how do you put a platform in place to harness that and then put a model, arrive at the data savings that you can showcase in terms of carbon footprint reduction. So again, to construct the algorithm, the model — I would say that was another challenging one, and quite a satisfying one, which we implemented for a large city installation for this lighting company.
Saurabh Gupta — President, HFS Research[04:40]
Both of them are fascinating, because it’s almost like the physical and the digital worlds are converging, right? And at the same time, it’s not just the physical and the digital world — we’ve been talking about IT and OT convergence. But I think there is a BT convergence, business transformation, that’s also happening, which was very evident in your first use case. Where is this world heading next, from your perspective? What are you looking forward to in the next couple of years? What are some of the big technology trends that you’re looking at?
Piyush Jain — Head of STX and AI, LTTS[05:29]
So IoT, like I said, has come a long way from those early days of just trying to connect and make the equipment smarter so that you can pull in data. The maturity has set in in terms of how the data is harnessed and structured, and now you can actually put analytics on top of it and perhaps even AI. The future trend is definitely trying to make it more AI-friendly. There are a lot of decisions that are still human-centric — most of the implementations still get the analytics piece into it, create those visualizations and dashboards, and still rely on humans to make the decisions depending on the trends that are being seen. But I think the future is really where the machines themselves will take decisions on when and what to do to prevent a certain incident from happening, or even predict the kind of correction that needs to happen, right? So let’s say in a plant, when these machines are generating data and the AI is actually monitoring, and they find out that there is a particular part that’s going to have wear and tear — today, the dashboard will give you that indication. It will go to the field monitoring person. They may take a decision on whether they need to order the part in advance, and so on. But the future would be where the AI can actually do the ordering by itself. When they know that there is a time that the wear and tear is going to happen, they can do a pre-order of the part so it can be there. The order can be delivered, and then obviously there may be a human intervention involved in replacing the part at the right point in time. But the kind of intelligence that is coming in because of the data and the advancements in AI, I think that’s going to make IoT extremely, extremely useful for business outcomes, and more importantly, that’s an exciting field that’s going to happen going forward. And again, generative AI is a new kid on the block. How do you bring in a more prompt-oriented, generative AI kind of intelligence into the IoT platform that you’ve already built, which can help not only look at what is being seen, but what can possibly be happening? What-if scenarios that could actually be done using that could be another area of exciting development. So a lot of possibilities are opening up, largely AI-driven, I would say, because on the data part, people have got a better understanding of it now. They have a better idea of how data needs to be harnessed and harmonized, but I think how you apply AI intelligently on top of it is where the world is moving.
Saurabh Gupta — President, HFS Research[07:55]
Yeah, I think we are moving towards a more and more autonomous world, as you rightly mentioned, where machines can make some decisions on their own and not everything has to flow through humans. But I think technology is moving faster than our culture and our mindset, and I think all the organizations have so much enterprise debt. While we’re talking about this autonomous world on one side, we are doing the same things the same way that we did 30, 40 years ago — maybe slightly cheaper, slightly faster, slightly better — and as a result, despite a lot of these technology promises, whether it’s IoT or AI, we’ve seen this story repeat again and again. The promise seems to never be fulfilled, right? There’s something or the other that’s holding us back, and it’s not technology. In your experience, what is holding us back to realize the full potential of whether it’s IoT or AI or, frankly, any other technology?
Piyush Jain — Head of STX and AI, LTTS[08:39]
In my view and my experience, technology advancements are definitely moving at a very fast pace, particularly in the last two to three years. The way AI has transformed itself is, I think, phenomenal. The challenge, obviously, is still the change management that is required at the human level. Now, let me go back to the case study I was talking about, this entire plant stuff that we did, right? The biggest challenge we had was when we got these models in place. The people who were actually on the plant monitoring the equipment, doing it the legacy way — when we gave them these dashboards, it took a long time for them to accept the dashboards, because they had to change the way they were going to use this new technology. So how do you make sure that they accept it? The reskilling that is required for them to be able to adapt to these new changes is a cumbersome process, and I think our customers themselves went through quite a painful journey to make sure that they’re able to get the people adapted to the new technology. Now, if you look at consumers in general, we are far more adapting to the technology with everything that we do, with our phones today or with our cars today and stuff like that. But when you go to the industrial world, I think we’re still behind in terms of adopting the technology. There could be fear of jobs. It could be fear of learning new methods. There’s a general fear that this technology is going to replace people. But my experience largely says it’s not replacing people, it’s becoming an assistant to people. People obviously will have to learn new ways of doing things, but I think the change management has been the toughest when it comes to adopting and getting the full potential out of this. On the consumer side, because consumer behavior has been changing much more rapidly, you don’t see it as often. But on the industrial side, where things are still in a lot of legacy and old ways of working, the change is quite dramatic for them. So that is, in my view, where the maximum planning, maximum effort from the management is required to ensure that this money that is spent—
Saurabh Gupta — President, HFS Research[10:40]
I completely agree with you. I think it’s the mindset, the culture, the change that sort of holds us back. It’s not really that we don’t have the technology or the toys to play with.
Piyush Jain — Head of STX and AI, LTTS[10:40]
Absolutely, absolutely.
Saurabh Gupta — President, HFS Research[11:10]
So, flipping that question — you’ve talked about how important it is for enterprises and clients who want to adopt these technologies, whether it’s IoT or AI. But I think our model as service providers will also need to change, right? And we’ve been doing… yes, we’ve been implementing new technologies, but our model of servicing has been the same. How is LTTS changing as you look at all these emerging technologies, AI? You know, we at HFS are talking more and more about services-as-software, not just software-as-a-service, and the whole servicification of practically everything that we’re talking about. So how are you guys changing yourselves?
Piyush Jain — Head of STX and AI, LTTS[11:54]
For the last 30, 40 years, I think broadly it still comes down to the base thing of what is the skill, the experience, and the capability of the engineer that you’re putting on a particular program. But I think progressively we are seeing a lot of change happening, even on the client side, where they are coming forward and telling us that they want us to come and solve a problem for them. Don’t come and just tell us what you’re capable of. From then on, I think, because the technology is shifting so rapidly, even our customers can’t call themselves masters anymore, like it used to be, let’s say, 10 years ago, where they were the masters and we were more the followers. Now I think it’s a field where they’re also learning. And in some cases, perhaps they are learning from the service providers, because service providers have a much wider perspective of technology application across different industries. The idea is not to go and prescribe anything to the customer, but rather go sit with them, understand what truly their problem is, and then come up with a solution which could include components of something that we already have done, the experience we have gained from maybe some other industry, the kind of body of work that we already have — how do we assimilate all of that, but make it a solution that is specific to their business, solving their specific problem, and not just come and say, OK, I have done this in the past, I can just come and replicate it for you. So I think that is a shift that we are doing at LTTS. We’re trying to say that we want to be coming and helping you solve the critical problems you have in your business.
Saurabh Gupta — President, HFS Research[12:59]
Oh, that’s fascinating, Piyush. I would love to keep going on with this conversation, but before I let you go, I’ll ask you one final question. If let’s say we grant you one wish that could come true — and let’s make it only one wish — what would that be?
Piyush Jain — Head of STX and AI, LTTS[13:35]
If there is one wish that you grant me, it would be a more open mindset, both at the customer end and at the provider end, to really come together and say, OK, this is how we’ll be able to take this forward and adopt the new technology for the benefit of the world at large — the climate change-related impact, or safety and security impact, and so on. And I think as service providers, my wish is also that we go to them with an open mind, saying that yes, we want to work together and help solve the most critical problems that are being faced in the world today. So I think if we are able to do this jointly together, I’m sure there’s enough and more satisfaction and business benefits that both the customers and us can gain together out of that.
Saurabh Gupta — President, HFS Research[13:59]
Oh, this has been fantastic, Piyush. Thanks for taking out the time. I truly agree that the lines are blurring — whether it’s the physical world and the digital world, whether it’s software and services, whether it’s technologies, the lines are blurring, and it’s becoming a whole converging blob of problem-solving, essentially. And I like the fact that we have to have an open mind, whether you are an enterprise or a service provider. Nobody knows everything right now, and I think if you keep an open mind and start to look at this converging world where the lines are blurring, there’s probably a greater chance of success. So on that note, Piyush, thanks for taking out the time. I really enjoyed the conversation. Thanks a lot.
Piyush Jain — Head of STX and AI, LTTS[14:49]
Thank you, Saurabh. My pleasure.
