Thomas Power

headshot_cropped.jpg

2140 FRB

2505 Hayward St

Ann Arbor, MI 48109

I’m a PhD student in Robotics at the University of Michigan. I work at the ARM lab, advised by Dmitry Berenson. I previously interned at Honda Research Institute USA. I received my MEng in mechanical engineering from Imperial College London in 2016.

I am interested in applying tools from planning, trajectory optimization, and machine learning to robot manipulation problems. During my PhD, I have developed algorithms using the planning-as-inference paradigm and deep generative models to accelerate trajectory optimization.

news

Jun 09, 2024 Our paper on optimizing diverse sets of constraint-satisfying trajectories was accepted to T-RO!
Jun 03, 2024 I successfully defended my Ph.D.!
Feb 05, 2024 Our paper on learning sampling distributions for sample-based MPC using Normalizing Flows was accepted to T-RO!
Dec 21, 2023 I am excited to be on the scientific committee for the Back to the Future: Robot Learning Going Probabilistic workshop at ICRA 2024
Oct 01, 2023 Our Differentiable Probabilistic Robotics workshop at IROS 2023 in Detroit was attended by over 200 registrants!

selected publications

  1. tro_flow.gif
    Learning a Generalizable Trajectory Sampling Distribution for Model Predictive Control
    Thomas Power, and Dmitry Berenson
    IEEE Transactions on Robotics, 2024
  2. csvto.gif
    Constrained Stein Variational Trajectory Optimization
    Thomas Power, and Dmitry Berenson
    IEEE Transactions on Robotics, 2024
  3. flowmppi_gif.gif
    Variational Inference MPC using Normalizing Flows and Out-of-Distribution Projection
    Thomas Power, and Dmitry Berenson
    In Robotics: Science and Systems , 2022
  4. lvspc.gif
    Keep It Simple: Data-Efficient Learning for Controlling Complex Systems With Simple Models
    Thomas Power, and Dmitry Berenson
    IEEE Robotics and Automation Letters, 2021
  5. dale_paper_fig.png
    Learning When to Trust a Dynamics Model for Planning in Reduced State Spaces
    Dale McConachie , Thomas Power, Peter Mitrano , and 1 more author
    IEEE Robotics and Automation Letters, 2020