Kishan Badrinath, a doctoral student in the Department of Electrical and Computer Engineering at Texas A&M University, was awarded the Nokia Bell Labs summer intern award for outstanding innovation during his internship this past summer.
During the internship, Badrinath focused on applying deep reinforcement learning algorithms, which are algorithms that learn how to achieve a goal over time, to autonomous drone navigation for avoiding obstacles and collisions using Microsoft AirSim.
“The Microsoft AirSim physics-based simulator provides access to the near real-world drone equipped with many sensors, such as the first-person image view, depth image, collision sense with an obstacle and much more realistic sensory information,” he said.
Over the summer, he was able to develop these algorithms to make a drone fly over small distances by avoiding one single obstacle, then scaled that to larger distances or areas with more obstacles.
“The final experimental setup was to get a reinforcement learning algorithm to maneuver the drone autonomously over long distances with many obstacles. I managed to get this during the last two weeks of the internship. This helped me reach the top 18 among a couple of hundred intern work showcases to Bell Labs’ top executives. Getting an intern award for outstanding innovation for my summer work via mail after a month of my internship ending was an unexpected bonus.”
Apart from research activities that he took part in, Badrinath also attended various activities, such as lunches and technical talks hosted by Bell Labs.
Badrinath started studying reinforcement learning in fall 2018 when he joined the electrical engineering Ph.D. program at Texas A&M under his advisor, Dr. Dileep Kalathil, assistant professor in the department. He completed his master’s degree at the Indian Institute of Science (IISc) in Bengaluru, India, where he was first introduced to the world of research.
Badrinath explained that he had an early introduction to online learning algorithms at IISc but that Dr. Dileep Kalathil’s research interests at the intersection of reinforcement learning and control theory sealed the deal for him to pursue his Ph.D. in this area.
"I am really happy that Kishan received this prestigious award from Nokia Bell Labs,” said Kalathil. “He is always seeking opportunities to broaden his knowledge boundaries by reading research articles and attending seminars. I am really excited about the important problems we are working on now and I look forward to continuing our collaborations in the coming semesters."