Welcome to H²T
The Chair of High Performance Humanoid Technologies (H²T) at the Institute for Anthropomatics and Robotics researches and develops humanoid robot technologies and systems that perform versatile tasks in the real world in interaction with humans. The research focuses on the mechano-informatics of humanoid robots, visual and haptic grasping and mobile manipulation, learning from human observation, modeling and analysis of human movements, active vision and haptic touch, as well as software and hardware architectures and system integration.
The H²T is headed by Professor Tamim Asfour, who has led the Humanoid Robot Research Group at the Humanoids and Intelligence Systems Lab (HIS) since 2003 and has been developing the ARMAR family of humanoid robots since 1998.
Scientific contributions of the H2T were nominated for all four awards of the 2020 IEEE-RAS International Conference on Humanoid Robots.
Congratulations to all nominees and the winners!
The European project IMAGINE demonstrated how to endow robots with the ability to understand their action effects in the context of the recycling of electromechanical devices such as computer hard drives (HDD). KIT developed a multi-function gripper that is able to perform a set of actions, such as pick, drop, lever, unscrew, etc. that are needed for disassembly tasks. The actions are learned from human demonstration and represented as via-point movement primitives allowing the adaptation to new situations. See article at CORDIS EU research results.To the article in German
Together with the team of Prof. Wolfram Burgard from University Freiburg we organized the workshop on "Semantic Representations for Robotics through Continuous Interaction and Incremental Learning" at the 2021 IEEE International Conference on Robotics and Automation (ICRA 2021). Detailed information about the workshop can be found at the workshops webpage.
We are pleased to present the results of the last camp in the form of a trailer. The event takes place regularly and in cooperation between the KIT research group H²T, the KIT Center Information - Systems - Technology (KCIST), and the ZML. Here, students can learn basic concepts of robotics and computer science, which they put into practice by programming the humanoid robot NAO.
For more information about our school activities please click hereTo the video
together.grow.shape – under this motto, key players from research, economy and society from all over Germany will meet on April 21, 2021 for the virtual results conference of the high-tech forum. The high-tech forum is the federal government's central innovation policy advisory body for the implementation of the high-tech strategy 2025.
H²T's research on humanoid robotics and artificial intelligence has also been selected for presentation in an online video. Don't miss it!
With its real-world lab "Robotic Artificial Intelligence", KIT will make artificial intelligence tangible through bidirectional exchange between society and research.
"The exchange of knowledge and experience goes in both directions in a real-world lab: Research and application come together at eye level, so that we can then develop precisely the technologies that people really need and want", says Tamim Asfour about the project.
The KIT, funded by the Carl-Zeiss Foundation, is researching a new generation of personalized humanoid robots. These helpers are to take over everyday household tasks in order to ensure the independence of elderly people. In addition, exoscellets are being developed that support the personal mobility of senior citizens and enable a targeted training of their motor and cognitive abilities.
For further information, please refer to the public press release of KIT and to the other projects supported by the Carl-Zeiss Foundation.
On October 8, 2020, Dr. Klein from the Baden-Württemberg Ministry of Economics, Labor and Housing visited the Chair of High Performance Humanoid Technologies (H²T). Together with Dr. Stefan Suwelack, Managing Director of Renumics GmbH, the progress of the joint MeshAI project was discussed, which investigates the automatic processing of CAD models using machine learning methods.
In the Science Camp Robotics, interested students learn how to program and control a robot themselves. Due to current restrictions, this year's Science Camp had to take place online. More information is available on Twitter and from the ZML.
You can find further news about our institute in our archive.