Online path planning using Reinforcement Learning - Enabler for versatile robot systems in public spaces and industry Enabler für wandlungsfähige Robotersysteme im öffentlichen Raum und in der Industrie: Onlinebahnplanung mittels Reinforcement Learning

Flexible path planning for autonomous robots is required in productions with a high degree of variability and short product life cycles. Reinforcement Learning (RL) offers a solution, as it enables robots to adapt dynamically to varying conditions and automate complex activities. The article explains the basics of online path planning using RL,presents a concept based on waste collection in public spaces,and discusses its transfer to industry.