This project aims to develop expertise for robot systems that use self-learning and other machine learning approaches so these systems can operate in dynamic, unstructured environments. This involves situations with many interactions between robots, humans, plants and animals. The focus of the project is on technology and the ethical and social aspects of robot innovation in real-world environments.
Societal and social values play an important role in the development of autonomous robots. The requirements imposed by the dynamic environment of plants, obstacles and animals must also be taken into account. As these aspects are included in the design process, the result will be a ‘value-sensitive design’. To ensure successful innovation, society and technology must change in parallel. This co-evolution also involves the preferences of the end users of the self-learning robots. Technical innovation is thus linked to societal innovation, during which we can contribute to social and business goals.
From the technology perspective, a priority is to develop robot systems using self-learning and other machine-learning approaches that are based on data-driven and data-hungry learning algorithms. Examples include deep reinforcement learning and methods that need a lot of sensor data to perform robot actions. Autonomous navigation uses algorithms based on services that exchange data.