The “Co-operation” Roboticist and his “Flying machines”!

Vijay Kumar (born April 12, 1962) is an Indian roboticist and UPS foundation professor in the School of Engineering & Applied Mechanics with secondary appointments in Computer and Information Science and Electrical and Systems Engineering at the University of Pennsylvania, and will become the new dean of Penn Engineering from July 1,2015. Prof. Kumar is known for his research in the control and coordination of multi-robot formations. He is an alumni of IIT Kanpur, graduated in the year of 1983. Prof. Kumar was awarded the Distinguished Alumnus Award last year by IIT Kanpur.

Vox Populi got the opportunity to interview him after the award ceremony and his lecture.

(Prof Vijay Kumar – Prof. Kumar:)

Vox: Good afternoon,Sir.
Prof. Kumar: Good Afternoon.

Vox: Sir, tell us about your involvements in the various technical activities while you were here.
Prof. Kumar: I didn’t do any of that .I remember, my activities were mostly musical. We played in a band and we had an audio club too, where we used to tinker around with things. We used to provide audio for all the big events and even make new instruments. We also had a lights club

Vox: And what did the lights club do?
Prof. Kumar: Nowadays everything is off-the-shelf but back then they had to be made. You can imagine it as having lots of lights and then wiring them up .It was a great way to earn practical knowledge. There was a HAM club as well as an electronics club that was pretty active. We didn’t have Techkriti back then but the cult fest was quite popular.There were hall days and we used to have Inter -hall cultural competition.It was called galaxy.

Vox: Something in particular you remember about them ?
Prof. Kumar: The bollywood talk line about the “Liquid Oxygen main duba dunga…..” – that was our line and they stole it.That actually went on to beget lines as “Is aadmi ko Pakad ke Liquid oxygen me daal do.. Oxygen isko marne nai degi aur liquid isko jeene nai dega (Dump this man in liquid oxygen. The liquid won’t let him survive, the oxygen won’t let him die)”
(Everyone laughs)

Vox: So when did you enter into the field of Robotics?
Prof. Kumar: Oh that is another story . There are a lot of random events in my life. I just picked things up as they came my way. While I was here, I wanted to work on heat transfer, solar energy, heat exchanger design etc. because I thought energy was the next big thing. But then when I went to Ohio State, I met Ken Waldron, who would later become my advisor. He sold me the field of robotics. It was an upcoming field with something new.

Vox: So you had no kinship with Robotics while you were an IIT student?
Prof. Kumar: There were two things I was interested in when I was in IIT. One was heat transfer and the other was aircraft. The whole idea that a computer can fly a plane i.e. autopilot was very interesting to me. In those days the autopilot were used only for inter-continental flights. I was very interested in it and so i took up Aerospace Engineering when I joined IIT. And then my Seniors told me there is no market in India for Aerospace Engineering. Due to this and partly because my father was in Mechanical Engineering I switched to Mechanical. You know back then you had much less information — no internet. So what you learnt, you learnt by talking to faculty mentors and by talking to senior students.

Vox: So how did your shift to Aerial Robotics and Swarm Robotics come about?
Prof. Kumar: That story is a concrete one. In my PhD thesis I worked on a walking machine and that walking machine had 18 joints. Back then we had Intel 8086 processor which had a clock speed of 0.5Mhz. So, you could not write a control software that could control more than one joint on a single board which meant that you had to have 18 Intel 8086 computers and a 19th one to just pass messages between them thus making them coordinate with each other. In effect you had eighteen independent joints swarming together to make the vehicle move. So I started thinking about it back then and then when I left Ohio State University and I went to Univ. of Pennsylvania, I couldn’t work on walking machines because they were expensive. So,I started building multi fingered grippers and multi arm manipulators. And again every arm is independent and you had to figure out how to make them cooperate. That actually went on to mobile robots swarming.

Vox: That’s a pretty solid way to experience the importance of swarm robotics. How did aerial robotics get introduced?
Prof. Kumar: Aerial robots came a long time later because the invention that made it all possible was the accelerometer and then if you go back and look at why the accelerometer became so inexpensive you, will find that it was because finally the MEMS technology became mature and car manufacturers in US decided to deploy airbags and then every car had an airbag and there were lots of MEMS accelerometer being sold. Suddenly, the price of MEMS accelerometer came down and thats what made small UAVs possible.

Vox: Ok,why do you think we are not developing new products in the field of robotics in India? I mean there is a lot of good research on robotics going on over here but not so much of indigenous products in the market.
Prof. Kumar: I personally think there is a lot of brain drain from the tech sector to non-technical sectors (someone said “financial”). When I came to the reunion ,I talked to the top three students on campus.It is true they were working in a tech field but it was clear they were interested in careers in completely different fields. I remember when I graduated people like me were criticized because IIT was training us and we were leaving the country to go to the US. But at least people like me stayed in the tech area, and contributing to education and advances in technology. Technology does not know geographical boundaries, and there should be incentives to recruit and retain smart young people in science and technology.
But you know now you have an innovation-incubation centre like SIDBI, I think things are changing and there are a lot of small startups and if one of them makes it big, suddenly everyone will want to create startups.

Prof. Venkatesh : Let me tell you what is actually happening. A lot of modern technology is being built from materials off-the-shelves. As you said I also believe in getting things off the shelves to build things.But in this way you are never going to build a good motor, even if you try . So there is really no success in this sector.

Prof. Kumar : That is not necessarily the only road to innovation. Innovation is mostly about novel ways of integrating components and subsystems.

Prof. Venkatesh: Yes say you start a company and you make something , you will need to sell at least 50,000 products to keep your company standing and that is very difficult in India. Here, the markets aren’t that big. We can’t sell robots, we aren’t competitive for that reason. These typical marks,I believe we can’t reach because the markets aren’t that big.

Prof. Kumar: Oh come one, India innovates in softwares and are really booming in it.

Prof. Venkatesh :Precisely,in things which don’t require material.

Prof. Kumar: That’s true even in the US. Remind me, I will send you the article about how “Hardware is the new software“. Software is something you can build in a garage or in somebody’s room. Hardware is becoming like that too. I believe in it. I think there will be startups in Robotics and if one of them makes it to success there will be a huge boom and by success I mean they have a good story and a high profile. They will have one soon and then Robotics will grow as a field to innovate in.

Interviewed by Amartya Sanyal