Excited to share our work, 'Traversing Supervisor Problem: An Approximately Optimal Approach to Multi-Robot Assistance', this upcoming Robotics: Science and Systems! Motivated by crop monitoring applications in agricultural robotics, robots can often get stuck in muddy or uneven terrain.

Very excited about this work with Ye-Ji Mun, Shuijing Liu, and Katie Driggs-Campbell! We found that a more complete map is not always a better map for RL. This lead to the proposed task-aware perception in the loop with RL, achieving safe crowd navigation in the presence of occlusions!