Robots & Me
"Never lose the child like wonder. It's just too important"
With my first ever robot in May 2009
Not a Humanoid :(
On a lazy afternoon in summer 2010, I was aimlessly flipping through the pages of a magazine when my eyes fell upon an advertisement for a robotics summer camp. Now for most kids, the word robot is analogous to a humanoid robot: intelligent machines that can fight monsters, transform into vehicles, eject missiles and many more things which have been glorified by cartoons and popular media. It was with these fantasies and great expectations in mind that I attended the summer camp, where they taught me how to build - a line follower. It was the first time in my life I had worked with electronics and I had certainly learned a lot, but quite frankly I was disappointed. The first thing I did when I came back home was tinker around with my line follower kit. I put together some bits and pieces to create a torso, head, and a few makeshift limbs, and I screwed them on my line follower. I still remember that moment when my modified line follower managed to move; my heart was swollen with pride for I had created my very first “humanoid”. In hindsight, I find the naivety of my 14-year-old self quite amusing, but I also believe it was a pivotal moment that changed my life forever.
Team Humanoid in September 2014 with our Biped
Undergraduate Robotics Team
I spent the next 3 years fiddling around with various electronic components. I would put together random parts and build the weirdest robots. I had no particular aim for the most part. The only thing I knew, was that it was fun. After school, I decided to pursue Electrical and Electronics Engineering as electronics had been an integral part of most of the robotics I had done so far and I believed it would give me the best foundation for a future in robotics. I remember turning up at my college’s robotics lab the very first day only to be turned down by the professor as the lab was only for grad students. I was disheartened but I soon found out that our college had many student technical teams that worked on underwater vehicles, solar cars, and even humanoids. In most Indian colleges, robotics teams are quite rare and humanoid teams even more so. So to have the opportunity to work on humanoid robots was for me, nothing short of destiny.
Why do bipeds fall ?
So we can learn and make better walking algorithms
Fall & Rise, Never Giveup
As a member of SRM Team Humanoid, I worked on various projects and participated in many competitions. But for the most part, it wasn’t easy. The team was quite new when I had joined. I remember in my first year, we only had 6 Dynamixel actuators which were not enough to build a biped, let alone a full-sized humanoid. But rather than the lack of hardware, the real difficulty was the lack of guidance. Although there were many professors in my college who was well versed with electronics, control theory etc. there was no one to guide us with humanoid robotics. Thus I mostly relied upon MOOCs, books and research papers to learn whatever I can. After joining the team, it didn’t take me long to realize that even something as simple as walking is an extremely complex task for humanoids. I spent weeks and months trying to get our biped to walk, only to see it stumble, fall and crash.
As the years passed, I learned a lot, our team grew and we got better hardware to work with. In my final year, I was appointed as the leader of my team. During this time, I spearheaded the development of three 20 DOF humanoid robots.
Shadow and Ghost- The twins
Charlie- 3D Printed Interactive Humanoid
Those tiny steps for the humanoid
A giant leap for me as a roboticist
I l led the team to represent India at Robogames 2017 and secured 4 medals (Gold for Penalty Kick, Silver for Biped Race, Silver for Freestyle, Bronze for Sumo Wrestling) in the humanoid league. My graduation project on Teleoperation of Humanoid Robots was selected as the Best Project in our department and was accepted for poster presentation at IROS.
But most importantly, I was finally able to design a robust walking algorithm, which for me, was the ultimate accomplishment.
My experience taught me one thing: When you want to be a Roboticist, you have to get very, very, very good at accepting failure, because building a robot, no matter what kind of robot it is, whether it’s a wheeled robot, remote-controlled robot, a walking robot, an artificial intelligence robot, you’re going to fail a lot. So the people who become good roboticists are the people who can accept that, pick up lessons from failure and move on. Finally, you triumph.
A resounding victory at Robogames 2017
Technology helps transcend disease and disability
I graduated from SRM University with a distinction and joined LV Prasad Eye Institute's Innovation Center as an LVP-MITra Fellow. LV Prasad Eye Institute is one of the best eye hospitals in India and they frequently collaborated with MIT Media Lab to come up with novel technological solutions for eye disease diagnosis. During my tenure, I was involved in many different projects but I primarily worked on developing a gaze tracking algorithm for Pediatric Perimeter, a novel device to quantify the visual field extent of infants.
I remember one particular day when a mother along with her 1-year-old child visited our lab to help us test my gaze tracking algorithm. The child had been declared blind at birth by doctors, so we expected no response to the visual stimuli produced by our device. But to everyone’s surprise, a few seconds later, my algorithm identified that the child’s gaze traced the exact path of the stimulus. After a few more rounds of testing, we figured out that the child isn’t actually blind, but has a motor disorder. I still can’t forget the mother’s reaction; she held her child in a tight embrace as tears of joy trickled down her face.
During the hustle and bustle of life, quite often, we fall into a routine of relentlessly trying to meet deadlines, finish projects, publish papers and so on. But its moments like these that remind me what we are truly out here to do: to positively impact the lives of people.
IIT Hyderabad Robotics Lab
After my fellowship, I came across Dr. R. Prasanth Kumar's work on extreme-legged locomotion at the Indian Institute of Technology, Hyderabad and found it very intriguing and well-aligned with my interests. So I joined his lab as a Research Assistant and worked here for 1.5 years.
To be very honest, I was flabbergasted at first on seeing the equipment at the Robotics Lab- Enough Dynamixel Pro Actuators to construct two full-sized humanoids, Hokuyo & Velodyne LiDARs, Nao Robots, different kinds of manipulators and state of the art machinery etc. I felt like a kid at an amusement park on my first day here.
This vast amount of resources opened a whole new world of possibilities to me. I worked on bipedal and quadrupedal locomotion, rehabilitation robotics, quadcopters, manipulator control and Computer Vision etc. and I have learned a lot in my time here.
I had a long-standing dream to build a full-sized humanoid robot and make it walk and I accomplished it at IIT Hyderabad.
Now, I move on to the Robotics Institute at Carnegie Mellon University to pursue my Master of Science in Robotics.