Robotics

The cluster develops autonomous robots and their interfaces with humans.

Robotics addresses major societal challenges. Securing and increasing welfare by continuously raising the productivity of work is the most obvious benefit of robotics. Robots, in their broader definition, will provide substantial support also in medicine and healthcare and will help address the problems of our aging society as well as of logistics and future transportation. The cluster primarily addresses the computer science aspects of robotics such as:

  • machine vision, e. g. for micro aerial vehicles and wheeled robots as well as for grasping and manipulation.
  • artificial intelligence and machine learning for planning and perception, e. g. for use in cameras that automatically detect faces, for robots which improve their grasping and manipulation capabilities over time as well as in online music streaming services or video platforms that analyze what you play in order to make recommendations based on what it thinks you like.
  • cognitive and physical human-robot interaction, e. g. for automated assembly without fences in industry. motion a task planning, that is used in robots which develop models of daily routines and their recurring changes, exploiting the long runtime to develop real spatio-temporal models of environments and activities.
  • autonomous navigation, e. g. in legged robots or quadrocopters that fly without human intervention, external sensors (such as GPS) or a-prior maps of the environment.
  • software architectures for robot control.
  • formal verification of autonomous systems software.
  • understanding the human central nervous system from the perspective of motion and manipulation and developing a new generation of robots based on these insights.

However, we also have strong expertise in robot design and control, neuroscience, and biomechanics.

We address robotics through an interdisciplinary approach and in direct collaboration with the departments of Mechanical Engineering, Electrical and Computer Engineering, and Medicine. 

Scientists

Current Activities

Gottfried Wilhelm Leibniz Prize (Prof. Daniel Cremers)

Prof. Daniel Cremers has been awarded the most important German science prize: the Leibniz Prize of the German Research Foundation (DFG). He is being recognized for his trend-setting research on mathematical image processing and pattern recognition. This research aims to endow machines with the ability to analyze and interpret visual data.

More information

ERC Consolidator Grant "3D-Reloaded" (Prof. Daniel Cremers)

The Grants awarded by the European Research Council (ERC) are among the continent's most prestigious research funding awards.

The ERC bestows the Consolidator Grants to outstanding scientists who have made a name for themselves through pioneering achievements between seven and twelve years after the completion of their doctorates.

Prof. Daniel Cremers carries out research in the fields of mathematical image processing and pattern recognition. He has been awarded an ERC Consolidator Grant for his “3D-Reloaded” project. The aim of the project is to create 3-D models of the real world from two-dimensional videos recorded using cameras, for example cell phone cameras. The resulting models could be used, for example, to calculate how much lignite was extracted from a mine over a particular period of time. They could also be used to explore the world’s most remote regions while sitting at a PC and could also enable film viewers to select the perspective from which they see an action. The analysis of the three-dimensional models is another of Prof. Cremers’ research interests. In medicine, this technology could enable the reconstruction of organs and the testing of the resulting models for anomalies. The models could also document and analyze the development of motion sequences performed by dancers and gymnasts.

Eligibility period: 01.09.2015 to 31.08.2020

More information

Selected Research Partners

The Human Brain Project HBP - Neurorobotics in the HBP

The field of Neurorobotics encompasses the intersection of computational neuroscience and robotics. The TUM led Neurorobotics subproject of the Human Brain Project is actively researching concepts within the field and developing the tools to allow researchers to fully explore simulated robotics driven by computational neuroscience models. Further, the development of biologically inspired, tendon driven robotics systems provides a unique research platform. These efforts allow researchers to explore the interesting space from computational neuroscience to intelligent robots and back.

The Human Brain Project (HBP) is a European Commission Future and Emerging Technologies Flagship Program intended to advance our understanding of the brain. It is designed to run for a period of 10 years with strong collaboration between 116 partnering organizations from all over Europe. With a total planned funding of roughly one billion Euros, the HBP is a large effort with very complex problems and goals. As such, the HBP effort is divided into 12 unique but interconnected research areas designed to foster interdisciplinary collaboration between research laboratories across Europe. Experts in the fields of neuroscience, biology, medicine, computation, robotics, and many more combine their efforts towards common goals. HBP researchers have published over 250 academic papers since the program’s inception in 2013 and many exciting research efforts are underway.

The Neurorobotics Subproject of the HBP is led by TUM and intends to place computational brain models within virtual robotic bodies that can interact within simulated environments. Imagine a computational brain model being able to sense – see or feel – the world around it and richly interact – through movement or manipulation. The goal of the subproject is help understand the concept of neurorobotics from computational neuroscience to intelligent robots and back. It primarily develops the Neurorobotics Platform, which offers scientists and developers all over the world the opportunity to connect brain models to robot models in various virtual environments. These neurorobotics experiments are executed in real-time on high performance supercomputing clusters through an easy to use web interface. This enables users around the globe to access otherwise unavailable performance from their browsers, even from mobile devices.

In addition to developing the publicly available platform, also contemporary topics such as biologically inspired robotics and learning are investigated. Much of this research focuses on the mouse – from fundamental biology to advanced simulations. Real mice are studied while their virtual counterparts are further developed. This complex and ambitious effort serves as an intermediary step towards the ultimate goal of understanding the human brain.

In March 2016, the NRP was officially released and the user base has been growing constantly. Researchers and the general public can request an account to access this platform. Only a few years after its launch in 2013, the HBP has come a long way. The TUM led Neurorobotics subproject has been progressing significantly, collaborating across subprojects, and generating results across its research and development efforts. With three separate chairs and the Fortiss institute being directly involved, it is also one of the biggest partners in its subproject.

Echord++ The European Coordination Hub for Open Robotics Development

The ECHORD project, funded by the European Commission, enabled bringing robotics technology from the lab to the market in more than 50 cases. The project also successfully tested the funding of sub-projects via open calls. Hence, this so-called cascading funding also plays an important role in the framework programme Horizon 2020.

Continuing in the tradition of the robotics research project ECHORD, the follow-up project ECHORD++ (The European Coordination Hub for Open Robotics Development, 2013 - 2018) is going even further in bringing innovation from lab to market. It promotes the interaction between robot manufacturers, researchers and users by implementing three different instruments: the Experiments, Public end-user Driven Technological Innovation (PDTI), and the Robotics Innovation Facilities (RIFs). With the Experiments and PDTI, ECHORD++ offers research consortia funding to develop robotics technology for real use-cases. The RIFs provide a unique chance to try out new business ideas and make field tests at zero risk.

These tools are tailor-made to meet the demand for innovative robotics technologies of the manufacturing industry, mainly Small and Medium Enterprises (SMEs) with small lot sizes and the need for highly flexible solutions, and public bodies, looking for robotics technology at competitive prices for tender processes.

Through the combined strength of all its unique instruments, ECHORD++ is probably the most application-oriented EU-funded robotics research project ever done.