Max Robotics — Global Service Robot Directory + US Certification

About

Jean Mercat is a leading researcher at the intersection of robotics, tactile perception, and safe autonomous systems. His work is defined by a commitment to bridging the gap between data-driven learning and real-world deployment, with major contributions in three key areas: large-scale robot manipulation, multi-modal tactile sensing, and risk-aware motion planning. Mercat is the primary force behind the **DROID dataset** (2024), a landmark contribution that has already garnered over 100 citations. This massive, in-the-wild dataset of robot manipulation is a critical stepping stone for training robust, generalizable robotic policies, directly advancing the capabilities of embodied AI systems like PALM-E and RT-2. Beyond vision and language, Mercat pioneers the integration of **tactile feedback** into multi-modal learning, arguing that contact-rich manipulation is incomplete without a sense of touch. His work on **Risk-Aware Prediction (RAP)** for planning addresses a fundamental safety challenge in autonomous driving: accurately estimating risk in long-tail, safety-critical scenarios where traditional sampling methods fail. By developing predictive models that explicitly account for uncertainty, Mercat is building the foundation for more cautious and reliable decision-making in interactive environments.