Learning Robotic Systems
Robots become part of our lives
Whether as a cleaning robot or as a portable application in rehabilitation after serious illnesses: Learning robotic systems can relieve humans in a variety of ways - even in environments that are dangerous or harmful to human health. Depending on the task, the requirements for the learning and adaptability as well as the controllability of the robotic systems vary. Questions on these topics are dealt with by WG 7 of Plattform Lernende Systeme.
Learning robotic systems should be understood as physical, technical systems in which AI is embedded. This enables them to operate in the physical world. They can be used efficiently and safely in complex environments for the benefit of people, society, and the economy - whether social environments where many people interact or environments that are dangerous or difficult for humans to reach.
Already today, robotic systems are increasingly being used in environments where this would have seemed inconceivable just a few years ago - for example, as body-worn systems in rehabilitation or to facilitate work in logistics, as cleaning or gardening robots in everyday life, or for customized production in small and medium-sized enterprises or in the skilled trades. Recent developments in the learning and adaptability of robot systems hold enormous potential. Particularly against the background of demographic change, changing conditions of globalization in times of pandemics, tense international relations as well as climate change, robotic systems with learning capabilities can make an important contribution to overcoming the associated challenges.
The potential of different forms of adaptive robotics systems for individual areas of application needs to be explored. The aim is to identify the prerequisites for the development of adaptive robotic systems and the challenges that need to be addressed in order to realize this potential. Associated with this are new business models made possible by learning capability, but also requirements for the controllability of learning and adaptability to enable safe and reliable use of the robotic systems.
Working group 7 headed by Ms Elsa Kirchner (University of Duisburg-Essen, DFKI) and Mr Jürgen Beyerer (Fraunhofer IOSB, KIT) focuses on these issues on the Plattform Lernende Systeme.