Invited Talk by Mr. Sanjeev Sharma on Self-Driving Technology for Indian Environments

Invited Talk by Mr. Sanjeev Sharma on Self-Driving Technology for Indian Environments

Title: Self-Driving Technology for Indian Environments
Speaker : Mr. Sanjeev Sharma, the CEO and founder of Swaayatt Robots
Host Faculty : Dr. Sparsh Mittal
Date: 20th June 2018
Time: 10:30am to 11:30am.
Venue: Auditorium


While there are several self-driving startups and companies around the world working on self-driving technology for structured environments, we at Swaayatt Robots are working on enabling self-driving in Indian unstructured environmental conditions with chaotic traffic. Typically, self-driving vehicles make use of various sensors, such as LiDARs, RADARs and cameras to generate the map of their surroundings, and detect obstacles and free space for navigation. At Swaayatt Robots, we use only off-the-shelf USB or IP cameras for perception. This is a very challenging problem, and requires making deep learning based algorithmic frameworks that are highly robust as well as run in real-time (typically > 35 FPS) on an entry level GPU.

At Swaayatt Robots, we have built navigation frameworks that enable our self-driving vehicle to (i) navigate at high speeds in cluttered environments, and (ii) navigate in previously unseen environments without having any detailed maps of the environments. This talk thus focuses on introducing the audience to the self-driving technology, and then discusses about Swaayatt Robots’ approach to enable self-driving in India, and the journey of Swaayatt Robots and the speaker.

A link to the demo of his self-driving car is here:

Speaker Bio:

Sanjeev has been researching in the field of motion planning and decision making (using reinforcement learning and convex optimization) since 2009. Between 2011 and 2013 he published 8 research papers in international conferences and journal. Since November 2014 he has been working full time on his startup, Swaayatt Robots (a bootstrapped startup), to enable self-driving in India. He has worked with prominent researchers, in the field of AI and Motion Planning, like Prof. Richard S. Sutton (his MS thesis adviser), Prof. Zvi Shiller, Prof. Matthew E. Taylor and Prof. Sridhar Mahadevan.

Sanjeev obtained his B. Tech in Electrical Engineering from IIT Roorkee in 2011. He then went to Israel for a research internship under Prof. Zvi Shiller (September 2011 – April 2012). He attended University of Alberta for MS in Computing Science, where his MS thesis adviser was Prof. Richard S. Sutton (2012-2014). He had been working remotely with Prof. Matthew E. Taylor since July 2011.

In 2014 he came back to India, declining PhD in Computer Science offer at UMass Amherst, to start his startup and work full time on enabling self-driving in chaotic Indian traffic.

Wednesday, June 20, 2018 - 10:30 to 11:30