Dr. Prakash Chakravarthi on Industrial IoT

Jeff Kavanaugh: Dr. Prakash Chakravarthi, your company connects machines in the plant floor to executives in the boardroom and work with dozens of companies around the world. Is there one whose story seems to especially resonate?

Dr. Prakash Chakravarthi: Yes, and in fact it's the most recent one that we have. We actually have an opportunity to automate thousands of gas wells in the mid-Atlantic area. And what's interesting, Jeff, is that historically the large oil and gas wells have been automated, but the low production wells, and I mean low production as really low production, these are wells that produce less than two barrels of oil a day, are monitored manually. People are sent once or twice a month to measure the readings and see if the tank is full because the tank is full, you're to send a truck, and if the tank is not full, you are to send somebody again. And it's a very expensive operation. So what the internet of things has done to the industrial markets is now the ability to bring the benefits of modern web services to low valued assets, and I think that's a very exciting thing to happen.

Jeff Kavanaugh: Industrial IoT and these assets are what we'll be exploring in today's conversation. Welcome to the Knowledge Institute podcast where we talk with thought leaders on business trends, deconstruct main ideas, and share their insights. I'm Jeff Kavanaugh, head of the Infosys Knowledge Institute, and today we're here with Dr. Prakash Chakravarthi.

Jeff Kavanaugh: Prakash is founder and CEO of Machfu, a company focused on the industrial internet of things or IoT, simplifying connectivity from edge to enterprise. He's also a successful serial entrepreneur. Prior to founding Machfu, Prakash founded multiple IoT and machine to machine businesses including Eka Systems, a smart grid connectivity company that was later acquired by Cooper Industries. Prakash's formal training includes a PhD in electrical engineering from Syracuse, an executive MBA from MIT, and bachelor's degrees in electrical engineering from Birla Institute of Technology and Science. Prakash, thanks for joining us.

Dr. Prakash Chakravarthi: Thank you for having me.

Jeff Kavanaugh: Well, excited to jump in on several fronts. Before we do, just to give some context, can you mention your journey or how you came to be CEO of Machfu?

Dr. Prakash Chakravarthi: My entrepreneurial journey began in 2001. At that time, a partner of mine who was actually in Japan in Toshiba research labs. He and I decided that we would found a company in the United States. He was very entrepreneurial. I had similar interests, and at that time we saw that the internet was becoming very big with people. We thought to ourselves,

Dr. Prakash Chakravarthi: And so we decided to found this company that focused on machine to machine communication. At the time we went into the automated meter reading markets primarily because energy was a big issue at the time. There was a lot of talk about the nation's energy needs being unmet, demand not being able to meet, exceeding supply and so on. But we are 10 years ahead of our time. I mean, the markets never really took off as predicted, but we still continued on, eventually raised a fair bit of capital, and sold Eka Systems to Cooper Industries, which is now part of Eaton Corporation.

Jeff Kavanaugh: Before we go any further. It's an interesting point. People think it's a compliment to be saying, "Oh, you're head of your time." But if you're 10 years ahead of your time, how do you hang in there, stay with it, because no one's there with you yet as far as a market?

Dr. Prakash Chakravarthi: It's very tough, Jeff, and I think we went through a lot of difficult times. And I think that's when I realized that when you talk about a moment of truth, that's when it dawns on you that when you go through some very difficult times, it's the people around you that make all the difference. I mean, the team that, the five of us when we founded Eka, we are still together after 20 years.

Jeff Kavanaugh: 20 years.

Dr. Prakash Chakravarthi: It's been a long journey together. We've been through a lot, many years of not taking enough money home, but, I mean, that's when you realize that if the right people are with you and you have the same goal and the same ideals, then I think magic can happen.

Jeff Kavanaugh: Once you get past that point, created a company, get it to a certain level, you sold it. What next?

Dr. Prakash Chakravarthi: Well, after Eka, we actually founded a consulting company, Captiva, because we still realized that it was still a little ahead. I mean, the perfect storm was not in place yet. What I mean by a perfect storm is that if you look at today, there is falling hardware prices, there's falling software prices. Cloud has made it really easy. There are many open source software available, many tools available in the market to create powerful solutions.

Dr. Prakash Chakravarthi: So what we did was we ran a consulting company and provided an outset for Fortune 500 customers. We worked with GE and several other companies, but eventually we got back into the products game because that's the DNA of the team. And the perfect storm, we realized, was now starting, and so that's when we jumped back into the Machfu journey, and it's been really exciting for the last five years.

Jeff Kavanaugh: The burning question I'm sure has to be, how did you come up with Machfu?

Dr. Prakash Chakravarthi: Yeah, that's an interesting story. We founded Machfu in 2015, and my daughter was 10 years old at the time, and we wanted to name the company. So I asked my daughter, and she asked me a very simple question, "What is it that you do?" So I told, we connect machines through phones. So she said, "Why don't we just name it Machipho, machine to phone, machine phone?"

Dr. Prakash Chakravarthi: So I brought it over to my CTO and said, "Hey, what do you think of Machipho?" He thought about it for some time and said, "Well, I could put a double play on it." He shortened it to Machfu. So he said, "Mach is speed and fu his skill," like in kung fu. So it had two meanings to it, and we thought it was a good name.

Jeff Kavanaugh: Well, a couple of points there. One, you probably owe your daughter a royalty of some kind. I'm sure at some point she'll ask for that.

Dr. Prakash Chakravarthi: Well, actually she's quite disappointed that we changed the name that she gave.

Jeff Kavanaugh: The other, it also shed some light into innovation and starting from scratch on something, how you can ricochet with ideas and where it can end up. And it is interesting as well because about speed, another part of kung fu, if you look at the Bruce Lee, "Be like water, my friends," and all that is that ability to be very responsive. And if you think about it, that's the role you're playing is you're connecting and helping those machines be responsive and connecting to the boardroom.

Dr. Prakash Chakravarthi: Yeah, and we had many years of experience in the industrial markets. I mean, we had worked across multiple verticals, utilities, industrials, rail, transportation, oil and gas, building automations, and so on. And we knew what the pressing problems were and why historically the market never scaled, what the barriers were, what the customer pain points are. And we thought if you could bring speed and skill into the play with the right combination of technologies, that would be a big contribution to the space.

Jeff Kavanaugh: For the benefit of some listeners, and more importantly, just to get your perspective on it, can you go into maybe a concept of industrial IoT? How is it different than the internet 1.0 now that you're talking about machines? And also for the corporate executive or director, what's special about it that's going to make the impact for enterprises?

Dr. Prakash Chakravarthi: That's a very good question, and it's very subtle. I mean, I get asked this a lot, and so IoT is there. I mean, I have Nest at home, Fitbit connects to my computer, so what's the big deal? It is a big deal and it's subtle and it's not easily recognized is the fact that if I tell you I can set the temperature of my home from New York, you get it, right? Very quickly. And how does it really happen, if you think about it? You have an app on the phone. Could be a temperature control app or it could be a banking app, you literally press the button, you open the app, you choose what you want. If you want to go to wire transfers, you go there. If you want to check your deposit, you go there. You type the numbers in. So I think in the consumer world, you don't realize how much a human being participates in the IoT to make it happen.

Dr. Prakash Chakravarthi: So I think the fundamental difference in the industrial side is that machines, they cannot participate by pressing things and choosing and configuring things the way they want. It all has to be done as though a human being is doing it. And that's a very complex operation, where there's complex algorithms, complex software, complex technologies. And that's why industrial IoT is a far more complex piece than the consumer side of things.

Dr. Prakash Chakravarthi: Now, but if you're able to do it, you get the same value. I mean, the value of modern web services in terms of efficiency, in terms of remote management, remote configuration, remote diagnostics, you don't have to go to the field. You don't have to send out a box truck to a distribution line to check a transformer, a cap bank, a fieldworker can look at it sitting in a McDonald's and they're having a coffee during a break.

Dr. Prakash Chakravarthi: I mean, the power of modern web services brought to an industry that's, I don't know, has $10 trillion in assets just in the United States alone across different verticals. So the value is tremendous in terms of efficiency, in terms of sustainability, and what can be done with the data.

Jeff Kavanaugh: Growing up on a farm and being at hundreds of manufacturing sites in my own career, what I find most glamorous is the unglamorous nature of what you're connecting to, and where these devices are often placed. Those three D's of dull, dangerous, and dirty, are where a lot of these remote operations are. Can you comment on how this maybe non-stereotypical, high-tech application actually plays out? Because usually you think of all the cool internet and AI types of things in these nice sterile office environments or a semiconductor. What's it like out there where things are a little messier?

Dr. Prakash Chakravarthi: Yeah, I mean, I can go back to our latest project. There is a well, say in Pennsylvania, that produces two barrels of oil a day. So at today's rates that's like $100 in revenues per day. So in a year, it's about $36,000 in debuts. Now, you still have to send somebody once a month monitor and empty the tanks and so on. So they spend roughly about $200 per visit. What that means is that 10% of the revenues are consumed just in operations, and that doesn't include maintenance and so on.

Dr. Prakash Chakravarthi: Now, if you're able to automate it with pressure sensors and tank level sensors and flow sensors and use the cloud as a backend, there's a tremendous savings in cost. And as unexciting as it may be, it gives a return on investment within the first two years.

Jeff Kavanaugh: Once again, you're listening to the Knowledge Institute where we talk with experts on business trends, deconstruct main ideas, and share their insights. We're here with Dr. Prakash Chakravarthi, CEO of Machfu, and expert in industrial IoT. As you're discussing this, a lot of it has to do with the oil patch or the sumps, remote assets. What are some other industries where you've seen application for this?

Dr. Prakash Chakravarthi: Well, I think it's all of that. We are young now, and we are focused on oil and gas markets, but we do have customers in the utility markets as well as in industrial automation. So in the utility markets there's a need for detecting faults. I mean, that's one of the big applications. If a fault on a line means that wattage of the current exceeds a certain bound, could be an upper bound or a lower bound. So even until recently as five years ago, utilities would actually send folks to a substation and there would be different crews that walk on each lateral feeder to see where the fault is.

Dr. Prakash Chakravarthi: And now there are new sensors that are come up that can do smart algorithms, figure out where the faults are. And automating all these things has a very big application because utilities set rates, the public utilities set rates depending upon various metrics, there's SAIDI and SAIFI [System Average Interruption Duration Index and System Average Interruption Frequency Index] and so on, which essentially led to how much outage is there, what is the frequency of the outage, what's the duration of the outage? And so on. It's a very big application and it can play a big role there in terms of reliable electricity.

Dr. Prakash Chakravarthi: And another example would be in the industrial side on factory floors, a tremendous number of rotating equipment, as you know, slabs, mortars and so on, and to predict their failures means that there's a limited cost to the companies that own and use them. So vibration sensors are used to monitor machine signatures. As the age and as they're closer to failure, they put out a certain signatures that they don't normally put out. And we are working with the customer to monitor the machines for failures, and that's a very big application as well.

Jeff Kavanaugh: If you're a director, vice president out there and you're excited about this, but you're having difficulty in getting it funded, what are some of the more exciting things that you can convey or maybe ideas you can give to someone to get these projects funded so they can start seeing these benefits?

Dr. Prakash Chakravarthi:: To be candid with you, that's always a challenge. I would say this. Look at all the benefits that web services is bringing to consumers. You don't even have to leave your home. I mean, you can order your furniture online. I mean, you can order your food online. You can even order a car online now. You don't even have to leave your home. I mean, that's how powerful modern web services is. And just imagine the level of efficiency and benefits it can bring to the industrial world.

Dr. Prakash Chakravarthi: I would tell them that it's very similar. Start with a pilot. And often I think some of the progressive ones do start with a pilot, but it doesn't go beyond that for a particular reason because a lot of them see the benefits, and then they talk to a company or they identify a group of companies and pick one and do a pilot.

Dr. Prakash Chakravarthi: But the reality is that things change really fast. I mean, if you look at how much not just the communications, but the world itself has changed in the last 20 years, you have to plan. Say I'm solving a certain problem today and a new problem comes, having deployed technologies that can adapt. For example, you may have a smart phone that you bought to talk, but then you want a banking app, all you do is download the banking app. So even though that was not the original intent, it's the same thing that you're using later to solve a different problem.

Dr. Prakash Chakravarthi: And I think many of these pilots fail because they deploy technologies that don't anticipate emerging problems and then they find that, oh, I’ve got to reroute all over again. So the first effort would be to convince them or have them believe that web services is a big benefit for the industry markets. The second effort would be to, once you want a pilot, pilot technologies that are expandable in the future.

Jeff Kavanaugh: You and I talked earlier, and while conventional wisdom glorifies analytics, analytics, analytics everywhere, you've expressed some reservations about blindly following it. Can you give some perspective on that?

Dr. Prakash Chakravarthi: Yeah, and I mentioned this to other folks before, it's like the joke that goes, "The marketing firm, after doing a lot of data analysis, concludes that everybody who buys a left shoe also buys a right shoe." So we actually have a practical example with this.

Dr. Prakash Chakravarthi: We were doing a project for a large utility in Asia, a water utility. And they supplied us with data on their water pipe breaks because they wanted us to predict future failures, they wanted to know where it would occur. So it's really common for people to, once the failure occurs, to go and fix them, but if you want to predict them in advance, it's a very hard problem.

Dr. Prakash Chakravarthi: So it turned out that in a few years’ time period, that many of the failures that occurred, occurred outside odd numbered homes. And to a human being, this would come across as a simple coincidence and they would just dismiss it. But if you feed it to a pure analytics platform, it would actually conclude that you need to put more resources in odd numbered homes because that's where the pipe bursts are likely to occur in the future.

Dr. Prakash Chakravarthi:: So I think that data analytics is a field that is going to grow in the future. I have no doubt about that. But you got to be very careful because it's very easy to come to the wrong results as often as you come up with the right results.

Jeff Kavanaugh: You spend a lot of time out in the field talking with people, both venture capitalist as well as clients, and investigating the future. What is a trend that's coming up that gets you most excited about the future in industrial IoT?

Dr. Prakash Chakravarthi: Well, industrial IoT is at a very early stage. I think people are just beginning to realize the benefits of connecting the device to the cloud. So we still have a very long way to go, and that itself is a really exciting fact that it started. There's so much a distance to cover. But if you look at some of the emerging trends that I would say no one is talking about today, it's really, the excitement is all about the value in connecting the device to the cloud.

Dr. Prakash Chakravarthi:: But the true power of the internet was we're they able to connect anything to anything, anybody to anybody. So when machines start talking to machines, so in other words, a temperature sensor can tell the AC unit, "Turn it up or turn it down," without having to go to the cloud and having somebody else decide what should be done, that's the direction in which it would eventually go to.

Jeff Kavanaugh: You're an executive listening, what are the three things you can provide them to get a project going, or it's stalled perhaps, to make it successful? What advice would you give somebody out there?

Dr. Prakash Chakravarthi: At this time, I would say that there is a lot of technology in place to do it right. I think there are benefits in many cases have been shown. The mistake that's often made is not so much that they don't deploy the solution to solve the problem they have in hand, but more often the big technologies that are not scalable and expandable. I would say that that's probably what I would tell the executives that if you can pick the right technology that's scalable and expandable.

Dr. Prakash Chakravarthi: What I mean by that is that once you've solved a certain problem, a new thing comes up, and very often the new thing, it's very hard to subsume them into the old solution. And that's where I think a lot of the companies get stuck. And I will tell the executives that I think the right combination of technologies are in place as standards. We've come a very long way. If you imagine the industrial world 30 or 40 years ago, I mean, it is all legacy. The protocols that we used to communicate to the machines created the internet, so they were never designed for either networking or for security. And those are the protocols that are still being used today.

Dr. Prakash Chakravarthi: And there are just so many of them. It's a very complex market. People often tell me that, "Hey, I have this machine that speaks Modbus and I have another machine that speaks Modbus, why can't they just talk to each other?" There are two Chinese speakers. I mean, one may speak Mandarin and the other may speak Cantonese and not be able to understand each other. And it's exactly the same for machines. I mean, it's a just having them communicate to each other is a very complex problem.

Dr. Prakash Chakravarthi: And many of those problems have been addressed today, and so I think the kind of revolution that happened in the consumer space and the enterprise world, a lot of IT technologies can be brought to bear in the industrial markets.

Jeff Kavanaugh: You mentioned in passing, cybersecurity is a big thing these days, a lot of justified concerns about it. And sometimes people think that IoT is a point of failure that often isn't addressed when people think about cybersecurity. What's your take on the best way to approach to make sure that you still are safe whenever you connect with these devices up?

Dr. Prakash Chakravarthi: Yeah, I think that the best way to do it is to bake it in instead of bolting it on. Very often, I think vendors and companies bolt on security after the fact and that always leaves a hole. But if you bake it in like what we do with the enterprise ID these days, I think you're far more secure. I think we have the mechanisms to do that today.

Jeff Kavanaugh: Taking a very Machfu specific approach, what is it that you've tried to do to ingrain it in your product and in your company that's different or distinct from maybe the way others have approached this theory of IoT?

Dr. Prakash Chakravarthi: So two things. One that we touched on just now, we baked in security right from the get-go. Security is baked into all aspects of our system. So in a typical communication system, there's a seven layer model where you go all the way from the physical layer to the application layer. So we implement security every layer of the model.

Dr. Prakash Chakravarthi: And the second approach is to have a sandboxed application platform. I mean, we've been very, very diligent about it, making the application platform very similar to your Android phone or iPhone where you can create new applications just by downloading them. It's as easy as that. Historically, as you know, the embedded world was a monolithic world. So if you had to change one aspect of something, you had to change everything about it. So we've created a sandbox paradigm, which makes it easy to bring about new applications.

Jeff Kavanaugh: That's important. Experimentation is important to be able to continue to try new things. Well, I know your time is important, and you'll need to get going here shortly. Who or what has been a major influence on you? What's guided you all these years?

Dr. Prakash Chakravarthi: Well, I would say two people played a big role in my life. My father is one of them. He was, of course, in India in a very socialist system, but I think he was an entrepreneur at heart, and I think I got his genes. He did a lot of innovative things in a socialist environment. In spite of it being an environment like that, he was very determined, and I learned a lot from that.

Dr. Prakash Chakravarthi: The other person has been my PhD advisor, Professor Don Weiner. Although I was in electrical engineering, what he really taught me was how to think as opposed to... I think he taught me how to think more than he taught me electrical engineering. I mean the process of starting from a basic set of facts and some assumptions, how to methodically work to conclusions is something that I've prized. I think that's the best thing a teacher can teach a student.

Jeff Kavanaugh: Yeah, amen. How can people find you online?

Dr. Prakash Chakravarthi: Well, we are on Twitter. It's @MachFuIoT. We have a LinkedIn page, and they can always reach me by emailing me at Prakash@MachFu.com.

Jeff Kavanaugh: You can find details on our show notes and transcripts at Infosys.com/IKI in our podcast section. Prakash, thank you for your time, and a very interesting discussion. Everyone you've been listening to the Knowledge Institute where we talk with experts on business trends, deconstruct main ideas, and share their insights. Thanks to our producer, Catherine Burdette and Dode Bigley and the entire Knowledge Institute team. Until next time, keep learning and keep sharing.