Learning AI for DeXtrous Robots (AIDX)
Learning AI for Dextrous Robots (AIDX) advances autonomous, dexterous robot manipulation for real-world tasks. It uses deep reinforcement learning based on first-principles simulations and achieves zero-shot transfer to hardware. The work is a joint research lab between TUM and DLR, applying methods on the DLR Agile Justin with DLR Hands II. Beyond research, it aims to commercialize AI methods and a next generation of highly dexterous hands for industrial and service robotics.
Learning AI for DeXtrous Robots (AIDX)
Learning AI for Dextrous Robots (AIDX) advances autonomous, dexterous robot manipulation for real-world tasks. It uses deep reinforcement learning based on first-principles simulations and achieves zero-shot transfer to hardware. The work is a joint research lab between TUM and DLR, applying methods on the DLR Agile Justin with DLR Hands II. Beyond research, it aims to commercialize AI methods and a next generation of highly dexterous hands for industrial and service robotics.