Virtual Training Platform for Robot Learning
A platform integrating Unity-based robotic simulation with machinle learning
In the scope of the AI4DI research project, we are developing simulation environments in which AI agents can be virtually-trained to control robots to execute industry-relevant manipulation tasks. To provide high level of realism in the simulated physics and rendering, we use high-end physics simulators and game development engines like Unity. In the video below, we demonstrate how the virtual environments can be used in highly-parallelizable simulation to train an AI agent for solving a robotic manipulation task using reinforcement learning.