GRASP Lab - UPenn

GRASP's Kumar Lab collaborated with researchers from MIT's Department of Aeronautics and Astronautics to develop a new algorithm to find optimized, safe flight paths for robots by mapping areas and planning collision-free trajectories.

Video: https://www.youtube.com/watch?v=Pvb-VPUdLvg&t=8s

EvMAPPER is a GRASP collaborative project that introduces the first orthomosaic approach for event cameras, enabling map generation even in challenging light conditions such as direct sunlight or after sunset.
Website: https://evmapper.fcladera.com/

GRASP's Kumar Lab developed a novel learning-based trajectory planning framework for quadrotors that optimizes their trajectory by changing the optimization problem into a quadratic programming problem.

Video: https://www.youtube.com/watch?v=tA02dJz9ux8&t=21s

Researchers at GRASP have developed M3ED, a multi-sensor event camera that provides high-quality, synchronized, and labeled data that can be used on multiple robotic platforms.
Website: https://m3ed.io/

The Perception, Action, and Learning Group has found that sensory information is rarely task-critical in many commonly studied task setups. Read more on their findings:
https://sites.google.com/view/vosi-robotics/