Max Robotics — Global Service Robot Directory + US Certification

About

Preston Culbertson is a robotics researcher whose work spans multi-agent coordination, optimization-based planning, safety-critical control, and neural scene representations. He has made substantial contributions to the challenge of enabling robots to operate reliably in complex, uncertain real-world environments. Among his most influential contributions is a decentralized adaptive control framework for collaborative manipulation (66 citations), which allows teams of robots to jointly manipulate payloads without inter-agent communication or prior knowledge of system parameters — a remarkable feat of coordination without explicit coordination. Complementing this, his work on mixed-integer convex programming (MICP) for robot planning introduced supervised learning strategies, through the CoCo framework (36 citations) and related methods, to dramatically accelerate solve times and make online optimization tractable for real-world robotics. Culbertson has also advanced safety-critical control under uncertainty, developing probabilistic formulations of Control Barrier Functions and Input-to-State Stability that provide formal safety guarantees despite stochastic disturbances and model error (30 and 10 citations, respectively). His CATNIPS framework (19 citations) further bridges neural scene representations and probabilistic collision avoidance, transforming Neural Radiance Fields into collision-probability-aware maps. Across more than 200 total citations, his research consistently bridges rigorous theory with practical robotic deployment.