Original Link :


Mumbai: Vijay Kumar ’s name is almost synonymous with drones, unmanned aerial vehicles (UAVs) and aerial swarm robotics. Born in Patna, Kumar spent his first 21 years in the country but was not born with a love for technology. Rather it was his mother who forced him to take the Indian Institute of Technology (IIT) entrance examinations. He graduated from IIT Kanpur—the result of what he terms “one of the smartest decisions” of his life—after which he left for the University of Ohio, and later started working as assistant professor at University of Pennsylvania where he has been working for the last 25 years. Kumar, 52, is currently a professor of engineering and applied sciences at the university. In an interview on Saturday on the sidelines of the TEDx India conference in Mumbai, he spoke about his offbeat, unconventional career, aerial swarm robotics, the future of robotics in India and his involvement with a couple of start-ups in the country. Edited excerpts:

How did India shape your love for technology?

Until I joined IIT Kanpur, I did not know what technology meant or what discovering things meant. In my years, we transitioned from the 11th (year school) system to the 10+2 system, so I had an option to take the IIT exam or stay in the 10+2 (system). As a kid you always want to postpone exams, plus my mother forced me to take the IIT exam. So I took the exam with literally no preparation, which is unimaginable today, and I got in. I chose Kanpur because it was the farthest away from home. These two decisions turned out to be the smartest decisions of my life and I have my mother to thank for it.

How did you spend your years at IIT Kanpur?

IIT gave me a solid, extraordinary foundation, a capability to understand and assimilate a lot of things. I think IIT taught me how to think, not to trust the spoken or written word, but to question everything and think outside the box. I was fascinated by auto pilots and flying and I started off wanting to be an aerospace engineer but people then said you won’t get a job as there is limited scope in the Indian aerospace industry, so I switched to mechanical engineering and then I thought about research as a career because I went through two internships—one in my third and one in my fourth year—and I quickly realized that I wasn’t doing anything new.

I was basically copying what other people had done, and my homework at IIT was more challenging than the projects in the internships. So I decided I would try something else and decided to get into research. That was another smart thing I did. There were a lot of professors who were mentors to me, and encouraged me to pursue my career in research.

What inspired to do research in swarm robots (group of collaborative robots that are programmed to work like ants and bees)?

When I went to University of Pennsylvania from the University of Ohio, I didn’t have money to buy a large unmanned vehicle, but we got two low-cost robot arms and tried to understand how to take these two robot arms and use them to pick up objects and control them. That was our two-robot system.

In those days, we needed three students—two to manage the robots and one to manage the interaction. So we started thinking of how to build multi-robot systems that are independent of students, because again we didn’t have a lot of money. So the multi-robot system manifested itself in ground robots, but aerial robots continued to be elusive. It’s only over the last 15 years or so that we’ve had technology that is mature enough to allow us to build low-cost aerial vehicles.

How difficult is it to build an aerial unmanned vehicle from, say, ground robots?

There is a difference. If you think about building a humanoid like ASIMO (the world’s most advanced humanoid robot, made by Honda), the big problem is balance. We have an even bigger problem because we have to do balance while we’re floating in the air, so it’s like defying gravity. But paradoxically it’s harder to balance on legs than it is to float in air because both legs are controlled independently, but the moment they both touch the ground, there’s coupling which makes it difficult to co-ordinate.

When did you begin working on the drones and unmanned aerial vehicles?

We were trying since 1998. We were the first ones. We built an eight-rotor aircraft supported by a balloon in 2000 so that if it crashed, it would crash on the balloon. The problem in those days was that the sensors you needed to sense acceleration and velocity were bulky and expensive. Now they have become inexpensive and they are available in every smartphone. So now you can build things, focus on the algorithms and the estimation to control, rather than focusing on the funding—that changed around 2005-2006. Our first aerial robot experiments were around 2008, and then our live robots took-off and that became a key focus area in our lab. But the fundamental idea of multi robot co-operation transcends aerial, underground, underwater.

How much do these robots cost?

Costs are decreasing, roughly at a rate of 30% per year—be it sensors, communication, transportation costs—in the last two decades and that bodes well for robotics. You can’t stop creating this technology just because it’s expensive. You just have to wait till costs go down.

How can these swarm robots be used, say, in India?

My grandfather was a farmer but I never liked agriculture, because I found it very boring. But 80% of the workforce is in agriculture…you have water shortage and food shortage in India, so it’s time we focused on agriculture. We have been talking to some companies here to incentivize people working in farms, and identifying solutions.

Are you talking with the Indian government for funding of robotics research here?

Not really. I am working with a couple of start-ups (as technology advisor) and I would like to help them establish the robotics industry and drone industry. One company is Autobirdz Systems Pvt Ltd, based in IIT Kanpur (provides aerospace solutions for sectors like agriculture, aerial inspection, security and surveillance), and the other is called Vayu. It is based in the US but has a Mumbai office, and is interested in drug and vaccine delivery in remote areas. Both these start-ups are looking for funding and I think that such start-ups will change the education system.

You may say I’m guilty of taking shop elsewhere, but at least I’m still in technology. The bigger problem India has is that the smartest people hardly stay in technology development, and that’s the bigger brain drain, given that India played such a pivotal role in the IT (information technology) revolution, admittedly in the back end. 80% of the value-added robotics is because of software, robotics is a burgeoning field, (yet) why are there no investments to ensure that the smartest students today are drawn to robotics (in India)?. I think Prime Minister (Narendra) Modi’s emphasis on ‘Make in India’ may not be the right thing. It should be ‘Design in India’ and ‘create in India’—you don’t care who makes it, you only care who creates something. I think ‘Make in India’ is a slogan of the 20th century and not of the 21st century.