J. Michael Harrison
Papers
9
Total Citations
124
H-Index
6
About
J. Michael Harrison is a leading researcher at the intersection of robotics, machine learning, and control theory, with a primary focus on enabling safe and efficient autonomous systems. His work centers on developing theoretically grounded frameworks for safe active learning, where robots must explore uncertain environments while rigorously satisfying safety constraints. Harrison’s most impactful contribution is his “Safe Active Dynamics Learning and Control” framework, which provides a practical and provably safe approach to sequential exploration-exploitation—a cornerstone for deploying robots in the real world. This work has already garnered 49 citations since its 2022 publication. Beyond safety, he has advanced robot motion planning by learning sampling distributions to improve efficiency, and pioneered the use of meta-learning to create priors for rapid online Bayesian regression. His research also extends to cloud robotics, where he developed learning-based network offloading policies to enable resource-constrained robots to leverage powerful deep neural networks. With a growing citation record and contributions spanning from theoretical meta-reinforcement learning to applied multi-agent planning, Harrison is shaping the future of adaptive, safe, and intelligent robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2Learning Sampling Distributions for Robot Motion Planning20 citations · 2018
- 3Meta-learning Priors for Efficient Online Bayesian Regression18 citations · 2020
- 4Network Offloading Policies for Cloud Robotics: A Learning-Based Approach15 citations · 2019
- 5
- 6
- 7
- 8
- 9Integrated structure-control design optimization of an unmanned quadrotor helicopter (UGH) for object grasping and manipulation2 citations · 2017