J. Andrew Bagnell

About

J. Andrew Bagnell is a leading researcher at the intersection of machine learning, robotics, and autonomous systems, with particular expertise in reinforcement learning, imitation learning, and motion planning. His highly influential 2013 survey on reinforcement learning in robotics — now exceeding 3,000 citations — remains a foundational reference for the field, articulating how RL frameworks can enable sophisticated robotic behaviors while robotics simultaneously challenges and advances learning theory. Bagnell is perhaps best known for co-developing CHOMP (Covariant Hamiltonian Optimization for Motion Planning), a landmark trajectory optimization approach that has reshaped how robots plan efficient, smooth motion through complex environments, garnering nearly 1,700 combined citations across two seminal papers. His pioneering work on maximum margin planning and imitation learning established rigorous algorithmic frameworks for robots to learn goal-directed behaviors from demonstration rather than hand-engineered programming — contributions further elaborated in his widely read treatise "An Algorithmic Perspective on Imitation Learning." Early work on autonomous helicopter control and later contributions to rough terrain navigation demonstrate his commitment to deploying principled learning algorithms on real robotic platforms. Across his career, Bagnell has helped define the theoretical and practical foundations upon which modern robot learning is built.

Research Focus

Key Achievements

Top Papers

1. 1
   Reinforcement learning in robotics: A survey

   3,055 citations · 2013
2. 2
   CHOMP: Gradient optimization techniques for efficient motion planning

   982 citations · 2009
3. 3
   CHOMP: Covariant Hamiltonian optimization for motion planning

   738 citations · 2013
4. 4
   Maximum margin planning

   639 citations · 2006
5. 5
   Planning-based prediction for pedestrians

   469 citations · 2009
6. 6
   An Algorithmic Perspective on Imitation Learning

   379 citations · 2018
7. 7
   An Algorithmic Perspective on Imitation Learning

   370 citations · 2018
8. 8
   Autonomous helicopter control using reinforcement learning policy search methods

   278 citations · 2002
9. 9
   Learning to search: Functional gradient techniques for imitation learning

   206 citations · 2009
10. 10
    Learning from Demonstration for Autonomous Navigation in Complex Unstructured Terrain

    143 citations · 2010

Key Collaborators

Contact & Links