The new TUM Center TUM ELI: First AI experimentation room for automatic knowledge generation

NEWS, Work, Robotics, Perception, Artificial Intelligence |

The Science Council has recommended the construction of a new research building at the Technical University of Munich (TUM). The TUM Center for Embodied Laboratory Intelligence (TUM ELI) is to be funded with 51 million euros. The research building is scheduled to go into operation in 2028. Questions for the Director Start-ups & Infrastructure of the Munich Institute of Robotics and Machine Intelligence (MIRMI) Prof. Eckehard Steinbach.

Prof. Steinbach, what is the TUM Center for Embodied Laboratory Intelligence (TUM ELI) supposed to do?

Currently, the processes in the experimental laboratory sciences are still little or not automated at all. Scientists spend one third of their time on laboratory experiments, i.e. rather working through tasks. The goal at TUM ELI is to automate the design and execution of experiments. An "ELI AI" communicates with the researcher, considers experiments and suggests how they should be continued or modified. This holds the potential for new solutions, such as new DNA drives for drug transport in the body. So on the one hand, it's about automating laboratory processes, but also about generating new knowledge through artificial intelligence. New structures and new mechanisms need to be developed for this.

The TUM ELI is to focus on the nano- and micro-world. Why?

The basic idea is that tiny robotic structures are both the subject of research and part of the solution. We are moving into very exciting dimensions here, which bring with them completely new challenges. For example, existing procedures and technologies for communication or information processing from the macro world cannot be transferred to the nano and micro world. Completely new solutions are needed here. The research tasks range from "intelligent connected assistants" to "cooperative fabrication of nano- and micro-machines" and "overarching communication and information processing". At TUM ELI, for example, tiny micromachines will be created in the future that are able to move around inside people and, for example, cut out and analyse tissue particles. Modern laboratory equipment such as a scanning electron microscope and a nanoprinter will be available for use by macrorobots in the future. Here we are thinking in particular of mobile platforms or cobots that relieve researchers of tasks such as introducing samples and microscoping them.

You are a member of the Board of Directors of MIRMI. What is the advantage for MIRMI and also beyond MIRMI?

MIRMI does not yet have suitable premises and equipment for automatic knowledge generation in the laboratory sciences. This experimental space is now being created with the TUM ELI. The ELI not only houses the experiments, it is the experiment. MIRMI will thus gain more Principle Investigators and cover an even wider range of topics. The TUM ELI is intended as a "shared facility" and is to develop into a centre for cutting-edge research and attract scientists from all over the world.

Fast Facts

The 10-year goal is to revolutionise the process of knowledge generation in experimental laboratory science using the example of nano- and micromachines by means of intelligent and cooperative multi-scale robotics. The central scientific question is how cross-scale analysis and synthesis processes can be massively accelerated by intelligent robotic laboratory assistants that independently plan, modify and evaluate experiments. The aim is for these robotic systems to develop maximum efficiency in conjunction with each other and with humans. Through cross-location cooperation and the interlinking of internationally leading scientists in the fields of robotics and machine learning (Prof. Sami Haddadin, Prof. Angela Schoellig), communication and computing (Prof. Holger Boche, Prof. Wolfgang Kellerer), perception and human-robot interaction (Prof. Sandra Hirche, Prof. Eckehard Steinbach) and nano- and micromachines (Prof. Berna Özkale Edelmann, Prof. Hendrik Dietz, Prof. Friedrich Simmel), the TUM ELI is developing an integrative, interdisciplinary working culture as a nucleus for a new quality of young scientists.

Funding: According to the recommendation for funding research buildings (2024), 51 million euros are earmarked for the TUM ELI. In addition to the TUM ELI, five other research buildings have been recommended for 2024, including the Center for AI-based Real-time Medical Diagnostics and Therapy (CARE-MED) in Erlangen. A final decision on funding will be made at the Gemeinsame Wissenschaftskonferenz von Bund und Ländern(GWK) at the end of June, but it is considered certain.

Construction and employment: Construction of the TUM ELI is scheduled to start in 2024 and be completed in 2028. About 70 to 80 people, including 50 to 60 researchers, are to be employed in the new building, which is to be constructed in Garching.

The Bavarian Ministry of Science has already reported on the recommendation of TUM ELI.