Max Robotics — Global Service Robot Directory + US Certification

About

Yunlong Song is a pioneering robotics researcher whose work sits at the dynamic intersection of autonomous flight, reinforcement learning, and optimal control. His research has fundamentally advanced the field of agile robotics, with a particular focus on autonomous drone racing as a challenging testbed for intelligent control systems. Song's most celebrated contribution demonstrated that reinforcement learning could surpass traditional optimal control methods in autonomous drone racing, a landmark result that garnered 194 citations and reshaped assumptions about machine learning in high-performance robotics. Building on this, his 2021 work on deep RL for drone racing (185 citations) tackled the notoriously difficult problem of time-optimal trajectory planning without requiring prior waypoint knowledge. His development of Agilicious, an open-source agile quadrotor platform (127 citations), has provided the broader research community with a standardized foundation for advancing perception, planning, and control research. Song also bridges the gap between model-free and model-based approaches, exemplified by his Policy Search for MPC framework (108 citations) and the Actor-Critic MPC architecture. His Flightmare simulator further democratized quadrotor research by offering flexible, high-fidelity simulation environments. Across domains spanning aerial robotics, autonomous racing, and ground navigation, Song's cumulative impact reflects a researcher consistently pushing autonomous systems to their performance limits.