Max Robotics — Global Service Robot Directory + US Certification

About

Shivin Dass is a roboticist at the forefront of scaling robot learning through data. His research centers on large-scale robot manipulation, teleoperation, and robust policy learning, with a focus on bridging the gap between controlled lab settings and real-world deployment. Dass is the lead author of **DROID**, a landmark dataset featuring 350 hours of in-the-wild manipulation data across 80+ environments, which has already garnered over 100 citations and is shaping the next generation of generalist robot policies. He also developed **PATO** (Policy Assisted TeleOperation), a system that dramatically accelerates data collection by enabling operators to control multiple robots simultaneously—a critical bottleneck in the field. Additionally, his work on **model-based runtime monitoring** integrates interactive imitation learning to detect and correct policy failures in high-stakes tasks, enhancing system reliability. Dass’s contributions are foundational for making robotic manipulation robust, scalable, and deployable outside the lab.