AI RESEARCH
Dynamic Neural Potential Field: Online Trajectory Optimization in the Presence of Moving Obstacles
arXiv CS.AI
•
ArXi:2410.06819v3 Announce Type: replace-cross Generalist robot policies must operate safely and reliably in everyday human environments such as homes, offices, and warehouses, where people and objects move unpredictably. We present Dynamic Neural Potential Field (NPField-GPT), a learning-enhanced model predictive control (MPC) framework that couples classical optimization with a Transformer-based predictor of footprint-aware repulsive potentials.