Odest Chadwicke Jenkins

About

Odest Chadwicke Jenkins is a prominent roboticist whose research spans robot learning, human-robot interaction, humanoid control, and open-source robotics infrastructure. A professor at the University of Michigan, Jenkins has made foundational contributions to how robots learn from human demonstration and how they can be made accessible to non-expert users. Among his most influential achievements is his work on ROS interoperability, particularly the Rosbridge protocol and Robot Web Tools ecosystem, which collectively have garnered nearly 300 citations and dramatically lowered barriers to cloud robotics and heterogeneous system integration. His early investigations into nonlinear dimensionality reduction—specifically the ST-Isomap algorithm (187 citations)—provided powerful tools for modeling spatiotemporal structure in motion data. Complementing this, his research on perceptual-motor primitives and performance-derived behavior vocabularies advanced the science of deriving reusable skill representations directly from human motion capture. Jenkins also reshaped methodological thinking in human-robot interaction through his "Oz of Wizard" framework (179 citations), advocating for rigorous evaluation of robot behavior independent of human simulation. His continued work on learning from demonstration, including incremental subtask learning and behavior tree acquisition, underscores a career-long commitment to making robots genuinely teachable by ordinary people.

Research Focus

Key Achievements

Top Papers

1. 1
   Rosbridge: ROS for Non-ROS Users

   201 citations · 2016
2. 2
   A spatio-temporal extension to Isomap nonlinear dimension reduction

   187 citations · 2004
3. 3
   The oz of wizard

   179 citations · 2009
4. 4
   Deriving action and behavior primitives from human motion data

   165 citations · 2003
5. 5
   Dogged Learning for Robots

   132 citations · 2007
6. 6
   PERFORMANCE-DERIVED BEHAVIOR VOCABULARIES: DATA-DRIVEN ACQUISITION OF SKILLS FROM MOTION

   89 citations · 2004
7. 7
   Incremental learning of subtasks from unsegmented demonstration

   87 citations · 2010
8. 8
   Robot Web Tools: Efficient messaging for cloud robotics

   85 citations · 2015
9. 9
   Learning Behavior Trees From Demonstration

   78 citations · 2019
10. 10
    Sparse incremental learning for interactive robot control policy estimation

    58 citations · 2008

Key Collaborators

Contact & Links