Self-Driving Car
Mapping & obstacle avoidance on a real autonomous platform · Robotics Research Center, IIIT Hyderabad (2021)
Self-Driving Car — Mapping & Obstacle Avoidance
Robotics Research Center, IIIT Hyderabad · 2020–2021
Earlier work from my undergraduate / M.Sc. at the Robotics Research Center (RRC), IIIT Hyderabad, under Prof. K. Madhava Krishna. I designed and deployed the mapping and obstacle-avoidance modules on a real autonomous self-driving car platform, built the ROS pipeline that fuses perception outputs into an occupancy map used by the planner, and ran closed-loop field tests. The three clips below are 2× speed with audio stripped.
Demonstrations
Autonomous driving (open field) — closed-loop mapping + planner driving the vehicle through an outdoor path.
Obstacle avoidance — local-planner reaction to detected obstacles in the fused occupancy map.
Drive-by-wire via Xbox controller — a fun early milestone: wiring the car up end-to-end so it could be driven over an Xbox-controller interface. Validated the low-level control stack before layering autonomy on top, and became the teleop + safety-override channel for later testing.
Related Publications
- DRACO: Weakly Supervised Dense Reconstruction And Canonicalization of Objects. IEEE ICRA 2021. Rahul Sajnani, AadilMehdi Sanchawala, Krishna Murthy Jatavallabhula, Srinath Sridhar, K. Madhava Krishna.
- Multi-Object Monocular SLAM for Dynamic Environments. IEEE Intelligent Vehicles Symposium (IV) 2020. Gokul B. Nair, Swapnil Daga, Rahul Sajnani, Anirudh Ramesh, Junaid Ahmed Ansari, K. Madhava Krishna. Developed an automated ego-to-metric scaling algorithm from interest points on the road to enable pose-graph optimization without requiring loop closure.