In the summer of 2015, I joined Auro Robotics as a Motion Planning Engineer during their Y Combinator batch, when Sam Altman was CEO. Working directly with the co-founders and initial team, I helped develop a driverless shuttle from scratch - a bold vision to revolutionize campus transportation. This was the early days of autonomous vehicles, and we were building something that seemed like science fiction at the time.
The challenge was immense: create a fully autonomous shuttle that could navigate complex campus environments safely and reliably. As the motion planning engineer, I developed ROS-based path planning algorithms that integrated GPS and LiDAR data to enable precise navigation. We worked in a small garage in Mountain View, iterating rapidly on hardware and software, often pulling all-nighters to meet YC's demanding demo day deadlines.
Our breakthrough came when we successfully demonstrated the shuttle navigating autonomously through Stanford's campus during YC Demo Day. The presentation to Sam Altman and the entire YC community secured $2.1M in funding, validating our vision. This experience taught me the power of rapid prototyping, the importance of scalable system design, and how cutting-edge technology can be brought to market through focused execution and relentless iteration.
Motion planning algorithm outputs by resim with real-world data
Sam Altman, Srinivas Reddy, Nalin Gupta, Jit Ray Chowdhury, 2015, Mountain View
YC Demo Day, 2015, Mountain View
Mail: info@ysong.dev
Location: Frankfurt, Germany
Website developed with Claude Code