Joshua A. Levine

Case Western Reserve University, Intel (United Kingdom)

Papers

Total Citations

H-Index

About

Joshua A. Levine is a leading researcher at the intersection of robotics, control theory, and artificial intelligence, with a primary focus on developing algorithms for autonomous systems. His most influential work, "Sampling-based planning, control and verification of hybrid systems" (2006, 75 citations), pioneered the adaptation of robotics motion planning techniques—specifically Rapidly-exploring Random Trees (RRTs)—to solve complex, non-linear control problems. This foundational contribution established a new paradigm for verifying and controlling hybrid systems, bridging a critical gap between robotics and formal verification. More recently, Levine has advanced the field of deep reinforcement learning for physical robots, as demonstrated in his 2019 work on "Customisable Control Policy Learning for Robotics" (13 citations). This research addresses the significant challenge of sample efficiency and policy transfer, enabling robots to learn adaptable control policies more effectively. His work is particularly notable for tackling the "reality gap"—the difficulty of deploying learning-based controllers on physical hardware. Levine’s contributions have shaped how modern autonomous systems are designed, verified, and deployed, making him a key figure in the ongoing integration of learning and control for real-world robotics.