Daniel Leidner wins ERC Starting Grant
- DLR researcher Daniel Leidner has been awarded the ERC Starting Grant from the European Research Council for his RECOVER.ME project.
- The research project aims to enable robots to autonomously detect and resolve their own hardware problems.
- Focus: Spaceflight, robotics, artificial intelligence
For his innovative RECOVER.ME space robotics project, Daniel Leidner from the DLR Institute of Robotics and Mechatronics has been awarded the ERC Starting Grant – a research grant from the European Commission. The project begins on 1 May 2024 and will be funded for five years. It aims to equip space robots with the ability to independently rectify hardware malfunctions, addressing an important problem in current space missions where hardware problems often require manual intervention and can lead to mission failure.
Robot self-care
RECOVER.ME combines elements from robotics, cognitive science and artificial intelligence. At the centre of the project is the idea of equipping robots with human-like problem-solving skills which should enable them to independently deal with serious hardware problems. For future space missions, in which complex robots are used in co-operation with astronauts, corresponding metacognitive skills are of key importance.
Current methods rely on pre-programmed responses to specific hardware problems. The RECOVER.ME project, on the other hand, focuses on integrating error handling directly into the robot's cognitive architecture. This involves developing a novel 'knowledge representation' that helps robots understand how different hardware malfunctions could affect their functionality. Correspondingly, a planning process for self-programming will be developed. A key feature of RECOVER.ME is that robots can monitor and adjust their programming in response to various hardware failures. Complex error information will be converted into a format that robots can process and understand, equipping them with flexible and effective problem-solving strategies. This could significantly enhance the resilience and efficiency of space robots while reducing the need for external intervention in the event of hardware faults.
Daniel Leidner has been working at the DLR Institute of Robotics and Mechatronics in Oberpfaffenhofen since 2011. In the Autonomy and Teleoperation Department, he leads the group for fault-tolerant autonomy architectures, to which the ERC grant is now affiliated. Since October 2023, he has been advising the federal government on developing of a strategy for AI-based robots.
The ERC Starting Grant
In the 2023 ERC Starting Grant call from the European Research Council, a budget of 628 million euros is being used to fund a total of 400 projects. The success rate for applicants is 14.8 percent. A total of 87 ERC Starting Grants have been awarded to German institutions. The target group for the Starting Grant includes promising early-career researchers (two to seven years after their doctorate) at the beginning of their independent careers.