First AI experimentation space for automatic knowledge generation

There is now little if any automation in experimental research processes. Researchers spend one third of their time actually performing their experiments in the lab. “The goal in TUM ELI will be to automate the design and execution of experiments,” explains Prof. Eckehard Steinbach, one of the co-developers of the concept: “An ‘ELI-AI’ system will communicate with the researcher, come up with experiments and suggest ways of continuing or modifying them,” says the Director Start-ups & Infrastructure at the Munich Institute of Robotics and Machine Intelligence (MIRMI) and head of the Chair of Media Technology at TUM. “The use of artificial intelligence harbors the potential for new solutions such as DNA-based structures to transport drugs within the body,” believes MIRMI Executive Director Prof. Sami Haddadin: “TUM ELI is designed to embody a robot-based lab-as-a-service concept and also to generate new knowledge through AI.” This will require the development of new structures and mechanisms.