Postdoctoral Researcher, University of Illinois Urbana Champaign
On the 2025–26 faculty job market in Computer Science, ECE, and Robotics.
Email: jmotes2@illinois.edu
Documents: CV | Research | Teaching
Profiles: Google Scholar | Parasol Lab
I design intelligent multi-robot and human-robot planning frameworks and hardware/software accelerations for real-world robotic deployments.
Multi-robot planning explodes in complexity as we add robots, degrees of freedom, and task structure. My work exploits the fact that only some regions and phases of a problem require tight coordination, using frameworks like DaSH and ARC to decompose problems and match each subproblem to the cheapest planner with enough intelligence. This yields orders-of-magnitude speedups and scales to large teams of manipulators and objects in realistic rearrangement and factory scenarios.
Learn more about Multi-Robot Planning →
Robots are shifting from executing fixed plans to co-planning with people in homes, labs, and factories. I build systems where humans, robots, and AI services share task and environment representations and iteratively refine plans through natural language and extended-reality interfaces like ERUPT. The goal is to support dialogue and joint decision-making, not one-shot commands.
Learn more about Human-Robot Co-Planning →
Many modern robotic systems are limited by how fast core planning primitives can run, not by what we could plan in principle. I work with theory and systems collaborators to accelerate these primitives using computational geometry, GPU parallelization, and serialized multi-agent search. This lets us redesign planning architectures so large-scale multi-robot systems respond in real time while still supporting the rich co-planning behaviors above.
Learn more about Hardware & Software Acceleration →
Lazy-DaSH demonstrates efficient multi-robot task and motion planning using a lazy approach to hypergraph-based planning, achieving faster planning times and improved scalability.