Thomas Power

University of Michigan–Ann Arbor

Papers

1

Total Citations

16

H-Index

1

About

Thomas Power is an emerging researcher specializing in robot motion planning, autonomous navigation, and learning-based control systems. His most notable work introduces a novel approach to Model Predictive Control (MPC) for collision-free robot navigation, where he leverages normalizing flows — a class of generative deep learning models — to learn adaptive trajectory sampling distributions conditioned on environmental context, start-goal configurations, and cost parameters. This contribution addresses a fundamental challenge in robotic planning: efficiently generating high-quality trajectory candidates in complex, dynamic environments without exhaustive computation. By embedding learned priors directly into the MPC sampling process, Power's framework bridges the gap between data-driven learning and classical optimization-based control, enabling robots to navigate more intelligently and reliably. Published in 2024, this work has already accumulated 16 citations, a strong early indicator of its relevance to the robotics and autonomous systems communities. Power represents a new generation of robotics researchers pushing the frontier of how machines perceive, plan, and act in the real world through the integration of modern machine learning with principled control theory.

Research Focus

Key Achievements

1
H-Index
1
Papers
16
Total Citations
16
Avg Citations/Paper
🏆 Most Cited Paper
Learning a Generalizable Trajectory Sampling Distribution for Model Predictive Control
16 citations · 2024
📈 Most Prolific Year: 2024 (1 Papers)
🤝 Key Collaborators: 1
🏛 Institutions: University of Michigan–Ann Arbor

Top Papers

  1. 1

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago