About

Sohan Rudra is a highly skilled Pre-Doctoral Researcher at Google with over 4 years of relevant experience in the field of autonomous driving. He is currently working as an Undergraduate Student Researcher for Team AGV, a research group based in IIT Kharagpur, with a vision to develop the first self-driving car in India, made for Indian conditions. Sohan has developed and implemented various state-of-the-art motion planning algorithms for autonomous driving and has worked on various high-precision sensors, including LIDARs, IMUs, and GPS, as well as industrial-grade robot prototypes and mid-sized vehicles for developing and deploying several state-estimation and planning algorithms. He has also worked on various Safe Reinforcement Learning algorithms for autonomous driving and has developed and maintained the full simulation stack for the autonomous driving system. Prior to his current role, Sohan worked as a Research Intern at Wadhwani AI, where he developed a Multi-View pipeline for the Visual Weighing Machine (VWM) project for reducing thickness ambiguity from predictions. He also formulated and implemented a novel Depth-Aware weak-perspective camera for Geometry-based Deep Learning pipelines and a model architecture fusing Multi-View with Geometry, which can make scale-aware predictions for a subject using a supplementary object. His overall algorithm increased the performance of the VWM pipeline by over 26% on validation and over 14% on test (MAE). Sohan has a Master of Science degree in Mathematics and Computing from the Indian Institute of Technology, Kharagpur, which he completed in 2021. He also holds a degree in Science with Mathematics and Computer Science from Rajendra Vidyalaya, which he completed in 2001. Sohan's tech stack includes Research Scientist, AI, Deep Learning, and test. He is a highly motivated researcher with a passion for developing cutting-edge technologies for autonomous driving. His expertise in motion planning algorithms, state estimation, and planning algorithms, and Safe Reinforcement Learning algorithms makes him a valuable asset to any team working in the field of autonomous driving.