Satish: Hello listeners, welcome to our brand-new episode of RawBotics. Today we bring to you a Carnegie Mellon alumnus who after graduating, straight away jumped into the chaos of Wall Street and learnt his roaps at BlackRock, eventually becoming an important pillar of the American multinational investment company. His love for robotics engineering eventually drove him back to India to start Niqo Robotics, giving up his American dream in the process to launch India’s first AI spot spraying robot in 2023. Please welcome Mr. Jaisimha Rao on RawBotics.

 

Jaisimha: Hi everybody.

 

Satish: So, Jai, investment banking, robotics and automation, agriculture, Carnegie Mellon, these are all very discrete things. Could you share with our listeners how did you traverse this journey? Yeah, I think in that list probably, Carnegie Mellon and the robotics engineering have the closest role. I went to Carnegie Mellon at a very interesting time. This was 2003 to 2007. There was this challenge called the DARPA challenge, which was funded by the American government. How do you build an autonomous robot? And Carnegie Mellon was the leader in that. So, I got bitten by the robotics engineering development work that was happening at CMU. And unfortunately, in 2007, instead of chasing technical wisdom, technical ambition, I chose money, and I landed up in Wall Street. So, (Wall Street was the thing at that time). 2007, all the smartest kids went to Wall Street. But there’s a big aspect because I think the kids were smarter than us and became Facebook employee number 10 and 12.  So yeah, I think in 2007, it was like the peak of the American economy. Wall Street was a good place for at least engineers like us to show off conjugative skills. So, I was one of them. I jumped right in. BlackRock Financial was an interesting company to work with. It was very strict. So, I clearly remember that we had to show up at 6.30 in the morning. (6.30). Yeah. And you show up at 6.31, it’s considered late. So, and every Friday they have a name and shame. So, during the week, whoever came beyond 6.30, they put it up on the big board and said, you guys came in late. So, I think when you’re 21, you need to discipline. The irony is I struggled to get people in at 10 o’clock here in Bangalore, which just makes me crazy. But sometimes you have to shift your requirements as you shift industries. So yeah, I just went to the flow. I got my computer engineering degree at CMU, got this good job. 2008 happened, financial crisis blew up. And because of that, BlackRock got additional business. So, this was in a group where the more chaotic it was, the more business we got to advise people how to manage the chaos. So, I stuck around for six and a half years. And then in 2014, I saw my boss and I saw the life that the boss was leading, which is you eventually move to Connecticut, you take the train, you vacation in the Bahamas, and that’s what life is. So, I think that freaked me up, the predictability of what is a career where I’ll be when I’m 30, 40, 50, 60, that is one. And I didn’t want to do that. So, I had my seven years at BlackRock and a bunch of savings during that seven years. And I decided to roll the dice and come back to India. And that time also, by the way, is this Shah Rukh Khan movie Swadesh was launched. (That inspired you). Of course, how can you not get inspired? (Very true). You leave and you want to come back to India and everything in India, a nation is rising, so you want to be a part of that. And I still believe that. So that is purely an emotional revolution. It’s a wild idea, but move to India. And I think why farming? You know, as a family, we have about 20 acres farm. It’s about five hours from the place we’re recording now. And that’s when I realised that farming generally is very wisdom-driven. So how do you make farming more precise, scientific? I think that was the inspiration of combining my technical masterminds probably four years at CMU and finding a use case to attack.

 

Satish: Very true. Farming has a lot of native intelligence involved in it. Yeah, definitely.

 

Jaisimha: Yeah, I mean, this place for wisdom, we’re not discounting that, but it can be 100% wisdom.

 

Satish: True. So can you tell us something about how, so eventually, Niqo Robotics started, you froze on farming, you decided you want to do something here and make some use case. So how did the entire Niqo Robotics journey came above? How did you move forward in this?

 

Jaisimha: Yeah, I knew I was ready for roller coaster, because we had a strap end, because it’s not been a straight-forward line for us.

 

Satish: And I think this was a time when robotics engineering was not the thing or very glamourous,

 

Jaisimha: Is it a thing now?

 

Satish: Yeah, yeah. (we still)? Definitely better than what it was, a decade back.

 

Jaisimha: Yeah, yeah. Now, definitely much better, but yeah, I think it’s 2015. I started the company. Interesting enough, I was allergic to Venture Capital. I said, I can do this by myself. I bootstrapped it from 2015 to 2019, and we started off as a drone company. So, how do you find drones, how do you fly drones over farmland and add value? So, you provide analytics. And over the four-year period, what we realised is farmers that’s been in farming community doesn’t like analytics, doesn’t like advice. They want a solution. So, the use case of drones, you fly over farmland, you would figure out where the plants are and tell exactly what is your population count. The farmers told us, hey, once you know what the population count is, come and fix it, wherever there is germination. And that the drones cannot do. So, it is a painful, four year product market-fit journey on trying to sell kind of advisory as a service to farming community and that didn’t work out. So, that’s when I also personally ran out of my bandwidth, and that’s when we realised we need to combine our analytics with a piece of hardware to deliver a solution, just not advise. So that’s when we got down to land-based robots that combined with analytics that we developed over four-years. And then from 2019 to now, yeah, I think we have a few products that we’ve developed specifically for farming. And I think the Eureka moment happened last year. So, just for folks who are listening, yeah, I started by 2015 and the Eureka moment happened in 2023. So, (eight years, eight long years). And you’re just chasing, chasing, chasing. And during that journey, you will get a lot of people say, oh, yeah, drones are amazing, it’s the future. So, robots are amazing, it’s the future. You’ll ask a customer – Is this what you want? They say, yes, you build that and they say, yeah, but I’m not gonna pay for it. So, I think you have to navigate all these signals from the market, what to build. And it took us, shouldn’t have taken us eight years, but it took us eight years. But last year, you know, our precision spraying robot, we’ve shipped 50 of them across India. And (that’s amazing), thanks. I think that’s a big number. I think globally also, when it comes to precision spraying technology and agriculture time in India, we’re probably right on top when it comes to the scale. But I think the most satisfying part is when we actually saw the tech in action. And the farmers saw that they could reduce 50% of their pesticide. And they opened the tank and they saw this 50% of pesticide and water left. And for that, the farmer is the real money. So, that is the Eureka moment. It took us only eight years. (But it definitely happened. That’s a positive. Yes, So, I think that’s something that we are proud of.

 

Satish: That’s great. So, you said that your robot helps in precision spraying, and then saves a lot of pesticide for farmers. (Yeah). So, how does technology enable this? How does the robot exactly do this if you could take our listeners to this?

 

Jaisimha: Yeah. Currently, you’re a farmer, you’ve have four acres, and you’re going to have a sprayer that sprays your entire field, what we call blanket spraying. So, irrespective of whether there’s a plant or not, whether there’s soil or not, then all the nozzles of the sprayer are on all the time. Which is silly, because the energy that you’ll draw is, if you’re a class teacher, you have 20 kids in your class, 3 of the kids are sick, you’re not going to give your entire class of pesticides or medicines. In farming, we’re doing exactly that. So, whether the technology claim, our technology puts a smart camera on a sprayer. So, we retrofit existing equipment. And, if you take a very easy example of pesticide spraying, we are spraying pesticide only on the plant and not on the soil. So, if you specifically again pick cotton, there’s almost four feet of gap between two rows of cotton, and two feet of gap between two cotton plants. (Okay). So, our technology sprays exactly on the foliage of the plant and doesn’t spray between the plants or between the rows. So, to a farmer, if you just take non, say, AI assisted spraying with blanket spraying, there’s an immediate 60% reduction in pesticide and water. And there are other consequences as well, because now you need less diesel on your sprayer, you can spray more acres per day. But that’s what our technology does. So, to answer your question, it is AI computer vision on the edge. We had to train on what a cotton plant looks like at six in the morning. It looks very different at noon, at five PM, it’s drooping. So, the soil looks different in Akola Maharashtra, vs Telangana. So, getting the dataset right, making sure you spray on every plant. I think those are all the technical orders that we have been able to cross.

 

Satish: Any challenges in terms of hardware, because agriculture, like, scale, would be very different across the country.

 

Jaisimha: Yeah, I think, in agriculture, most of the equipment is made for abuse. And as long as you are dealing with machinery, iron, steel, I think that those kinds of materials can take abuse. But as soon as you get into our world, now suddenly you have an media processor, you have PCBs, you have a camera, with a lens. So, heat, dirt, dust, some things, I think are here and there. I think we’ll get to this when we speak about this. But the need for excellent mechanical protection is probably a big blocker in robotics engineering. Because in general, you can try to combine, say, computer vision, that will work. But how does it work in the real environment? You need a good foundation. So, that’s been a big technical analysis challenge that we have been able to overcome.

 

Satish: That’s great. So, you said that you have already deployed robots in India. Do you also see a global market for your spraying robot as well?

 

Jaisimha: Yeah, the wonderful thing about agriculture is, every place in the world requires food. And to make food, there is probably some pocket of agriculture in every country. So, if there’s a global plan, what we are doing, there’s always an opportunity to move agriculture, pesticides, spraying from a blanket spray to a precision spray. And we are keenly looking at expanding global leadership. So, we’re excited about that. I think time in India, we do have home called advantage Now there’s 20 million hectares of agriculture in India. So, the scale of India is no country can match at all. So, we do need to leverage what we have going for us in India. The margins are a little bit tighter, but we do have to get along the India. But maybe opportunistically over the other way will go shorter.

 

Satish: Very true. So, since you mentioned that you have deployed the robots, there is something that I am very interested to know. So, a farmer typically might not be very aware about the robots since you’re talking four acres as the average holding. (Yeah) So, how does the farmer interact with the robot? How does that interaction happen? What is the kind of interface that you have given?

 

Jaisimha: Yeah, I think specifically for agriculture in India, if you look at it, it’s a very peak and troughs cycle. All farming and curry season happens within a few hundred days. And the rest of the season, you’re not doing much. So, we didn’t want to put the burden of technical mastermind adoption on the farmer. (okay) So, the 50 robust agricultural norms this last season, and we continue to operate, is it’s the farming as a service. It’s a robot as a service. So, we have trained technicians who are interfacing with the display on the sprayer. They are selecting on a screen which deep learning model. Is it a cotton deep learning model, is a chilli deep learning model. Then there are few variables there you can play with. When does spray start before the plant, when does it stop after the plant. So, that interface, the operator sprayer does. (okay) So, he in turn is the key man for adoption. The farmer himself or herself is just engaging the robot as a service. So, they engage us and say, yeah, four acres can you come and do precision spraying on us. So, if you ask me the operator, you won’t believe, given the experience with operating smart phones. In some cases, they are better at navigating the screen, selecting the model, selecting the specs, better than our own staff because they have done it more. I think that’s another important piece to robotics engineering. How do you make it super easy to use so that anybody can use a smart phone can. So, that’s been our go to market for India. And I don’t think we have had any technical analysis issues in terms of folks not understanding. The second part of the question, I also say, I mean, robot is the scary name? When you say robotics and automation people think, all kinds of things. In your case, it’s warehouse robots that are moving in a factory and trying to move things. In our case, if you are going on a road and you are seeing this sprayer spray, you won’t notice anything different. So, it’s the ability to turn, nozzles on and off. But to get that at a very precise level, you need the ability to see, the ability to process, and then finally the ability to act. So, the farmer, as soon as they see a sprayer turn on and off and spray only on the plant, they’re sold. So, we usually do a couple of lines, pro bono just to show that it works. And I think most of the time when they see it spray on your plant, they say, pura kardo. So, they don’t need to know are we using YOLO deep model or Media z SEN. We don’t really complicate that. You are spraying on your plant, you’re not missing any plant. Now go ahead and do this.

 

Satish: So, I think what would be the frequency of the spray, typically in a growing cycle?

 

Jaisimha: There’s about 7 sprays in a full cotton season. So, it’s quite sticky when it comes to how you’re interacting with farmers.

 

Satish: And since you mentioned that yeah when we deploy a robot in warehouse we have a very good infrastructure in terms of connectivity and internet. So, since you mentioned you are using edge, how did you solve this problem of connectivity or data? Did 5G help you in some way of what was your experience around this?

 

Jaisimha: I think for us, connectivity, we assume that there is going to be zero connectivity. So, (okay) it’s not only India in fact, maybe India is probably better connected rurally than the other parts of the world. So, more than connectivity, our technical challenge in outdoor robotics engineering is you will never know which farm you go to. Every farm is different, different undulations, different crops. There will be no north-star per say. There is nothing to say that this is zero zero. Yeah, correct. So, real technical analysis challenge that we have been at it is how do you make sure with different AI techniques, localisation, SLAM? How do you make sure when the robot moves on a millimetre or centimetre level, that reading is as pure as it is without noise. That, I would say, is more of our core challenge than connectivity. (Okay). And we are super jealous of you because in warehouse robotics and automation, all the techniques, all the literature out there is localisation of the robot in my warehouse. In an indoor environment. (Very true). And my robotics engineers don’t have the luxury. So, we’ll be in Akola one day, we’ll be in Punjab other day. And we still need centimetre level decisions. But you have been able to solve it to a large extent.) Yes, yes, we have. I think that’s the, when people think about robotics and generally the AI waves that we are currently on, you start peeling the onion, the core of what makes a robot robot is probably not an AI. Yeah. It’s these different deeply technical inventions that you need to implement, like, odometry, localisation, thinking about optimisation on your processor. But as hungry entrepreneurs, we will bundle that up all into the AI and sell that. But, since your audience is probably more robotics and automation focused, that’s where the real action is for us.

 

Satish: True. Very true. Since there is a lot of focus on green sustainability, we’ve been hearing a lot about climate farming. So, if you would talk to our listeners about climate farming, how does your robot fit into the biggest scheme of climate farming?

 

Jaisimha: Yeah, if you pick one of our farmer testimonials, before our technology, the farmer was spraying about 200 litres of mix of pesticide and water. And so, as soon as they used CN spray technology, spot spray technology, that 200 litres dropped down to 100 litres, (oh) 50% savings direct. So, with this number, and again, there are certain farmers who have experienced 58%, but just for a sake of argument, let’s assume a clean 50%. If we do a 50% savings, and you’re considering 100 litres of water on one spray, and multiply that across 33 million acres, just imagine the amount of excess pesticide and water and diesel that is currently being used for complete waste. I think this is not adding to the productivity, it’s just going into the soil, it’s contaminating the water. (Yeah, it’s probably having a lot of damage in fact). Yeah, so, if you start looking at agriculture in India, the areas where you’re going to be able to get the water, the areas where we can make a positive difference, this is the no-brainer invention. Where 200 litres versus 100 litres. And there is no effect on output. So, why aren’t we doing that? So, that is what we are excited about. It’s easy for us to sell to the farmer, the farmer sees it, it’s easy to understand and also nozzle going on and off. So, 33 million acres, 100 litres, this is not a climate storey, I’m not sure what is.

 

Satish: That’s talk about impact and that’s really it. (Yeah,) it’s really true. So, you mentioned that you got to work on DARPA projects when you were in CMU. (I was by stander in Hawthorne, but I did see it) Yes, you did see it. And I’m sure you would have seen a lot of in Carnegie Mellon as well. So, beyond agriculture in India. So, what are other robotic applications or robots that have captured your imagination? I’m asking this question because since you’ve almost been at the land when a lot of things were happening in US.

 

Jaisimha: I think autonomous driving is something that gives you goosebumps. We were in California a few months back and we borrowed my brother-in-law’s Tesla. And we just put it into autonomous mode, and the first time, user experience to see a car change lane, accelerate, decelerate, exit a road. I think the future is here. You can imagine, a fleet of this autonomous cars, doing… ..dropping under caves, doing chores. I think that’s my experience to it. So, I’m excited about it. Now, will we get to see that at scale, I think the last 5% is going to kill in the deck. Okay. But because we’ll obviously have this moral dilemma of, (yeah), what if an autonomous car crashes verses human car? (So, we have seen this playing out). Yeah, we’re seeing it in there. But I think that’s, like, a goosebump moment for 1. Number 2, something closer to what we are doing. In agriculture, again, it’s just to set context. If you look at a farmer’s life, the farmer buys whatever he or she needs to do for the season up front. So, they buy seeds, they prepare their land and the tractor, they will buy pesticide, and they buy all that at a retail price. So, they do all this effort. 180 days. I dip the soil, I put the seeds, I spray, I nurture it. And finally, when it comes to harvest, literally you have money on your farm. Whether it’s cotton, whether it’s fruits, vegetables. Now, that peak demand, essentially you certainly find yourself in farming. There’s a 3-day window where all of India, all of Maharashtra needs to take the cotton after cotton. It’s true. And you don’t have enough people to do that. So, you get in the ugly situations where only for harvest, the prices peak, like 3x 4x. And the irony is, once harvest is done, the farmer sells at a wholesale price. He’s buying a retail, selling at wholesale, and when he sees money on the farm, he’s at the mercy of marketers. If I had unlimited resources, and if I had relatively long, kind of 10-year plan, I would definitely try, I think, picking, how rested, picking robots agriculture is another area that I think is a big need for it. And it’s solvable. And it’s a huge problem. And the irony is, we have 1.4 billion people and send us a labour problem in India, which people can’t access, imagine other countries. And that’s number two. Number three is something similar to what you guys do. I think entering, I’ve been to a couple of warehouse robotics, fully automated robotics. I think it’s very, very gratifying to see small machines talk to each other and coordinate with each other. (This is what we call robotic symphony) yeah, it’s a good way to put it. So, I think it’s cleaner, it’s safer, it’s more optimised. So, I think warehouse robotics and automation, we need more of it. And I think the warehouse of dealing with so many packages is quite dangerous, they’re all grown very costly. So, I think Addverb and the work you’re doing is, I think you should do demo things. You should bring people to see one of your automated processes.

 

Satish: For sure, we have now put up a manufacturing facility in Greater Noida, which is one of the largest, where we have all these robots and automation systems that play for customers, where they can come and experience how the robotics and automation facilities are. So, just to end it on the lighter note, so if you have to able to give a become a robot for a day, what would you like to do?

 

Jaisimha: It’s always childhood dream of mine to become a fast bowler. (Oh) So, I don’t know. (While walking, do you still act?) Yeah, now I shifted the golf. (oh,okay) But yeah, I mean, if you see me in elevator here and the door closes, I’m practising my swing. (Yeah). By myself, So, we all do. But maybe it’s like one of the things where we should get up early in the morning, watch India tour Australia and see all these fast bowlers. If I can have a retrofit kit on me and make my 38-year-old body bowl at 160 kilometres per hour, like one very personal shallow bowler. (You can probably have an exoskeleton and we can help you with one). Ah, really? You want to make my dream come true? (Yeah, We have an exoskeleton). Okay. Okay. Hook it up to me and then I can show it in front of my family that I can bowl at 160. I don’t think I have ever crossed 90 kilometres per hour. So, I think that will be one shallow boyhood dream that a robot will help me accomplish. Yeah. (That’s an ambition that should definitely be fulfilled). I’m sure their other important problems to figure out in this world. But yeah, I think this is something personal for me.

 

Satish: Thank you. Thank you, Jai, for a very enlightening conversation. Thank you for your time.

 

Jaisimha: No, thanks for having this. I think, you know, we don’t get our fair share of limelight. The work you’re doing and the work we are doing. Because mainly because it does take time. The gestation period is so high. By the time you figure out the tech stack and go to market, you know, the world has changed. (Yeah). But I feel like the next 10 years is probably deep tech combining software with hardware. And we’re going to have multiple successes here. So, I think our time is coming. that’s the field I’m getting. I don’t know about you.

 

Satish: Yeah, for sure. Thank you.