Skill Development Workflow for a 6-DOF Robotic Arm to Reach Targets Using Deep Reinforcement Learning

Mar 4, 2024

The University of Deusto has been exploring and designing different strategies for autonomous learning of robotic skills (using Deep Reinforcement Learning) within the ACROBA project, from simplified mathematical models of the environment to speed up the process at the initial stages, to the use of high-quality Virtual Gym simulated scenarios designed for ACROBA use cases.

Check the video below to gain insights about the different steps of the process.