Sebastian Castro

Cornell University, MathWorks (United States)

Papers

Total Citations

H-Index

About

Sebastian Castro is a researcher at the intersection of artificial intelligence and robotics, with a primary focus on hierarchical planning, task-level abstraction, and autonomous control. His work addresses the fundamental challenge of enabling robots to reason about and execute long-horizon, complex manipulation tasks. Castro’s most notable contribution is his pioneering approach to active learning for abstract plan feasibility, which allows a system to efficiently predict whether a high-level plan can be physically realized before committing to costly low-level motion planning. This work, published in 2021, has garnered 13 citations and is critical for scaling robot autonomy in unstructured environments. Earlier, Castro made significant strides in modular robotics, developing methods to generate provably correct control from high-level tasks expressed in structured English—a contribution that bridges natural language and robot motion. His 2011 paper on this topic, with 12 citations, remains influential in the field of reconfigurable systems. Castro has also advanced the RoboCup Rescue Simulation framework by integrating modern AI algorithms, enhancing multi-agent coordination in disaster scenarios. His research is distinguished by its focus on practical, verifiable autonomy, making him a key figure in the evolution of intelligent robotic systems.