ThurAI (Thuringian AI) - Subproject RobinCare - Robotik in der Pflege



Building on the activities of the TZLR (Thüringer Zentrum für Lernende Systeme und Robotik), 12 pilot projects with a total funding volume of 2.56 million euros are being initiated within the ThurAI project (Thuringian AI) on the application and transfer of AI methods in practice.

To this end, the focus is on the following three key topics, which are of particular importance for the Thuringian economy and science:

  • Production and quality assurance by supporting the automation of production and manufacturing processes in industry, but also process monitoring and quality assurance.
  • Healthcare and medical technology through the use of AI methods in medical technology devices and in care facilities,
  • Smart City through intelligent, data-driven solutions for efficiency, citizen-friendliness and sustainability of modern municipalities.

Demonstrators are to be developed for these guiding themes within the framework of pilot projects, which will show the potential of intelligent or learning systems and illustrate application examples.

The RobinCare - Robotik in der Pflege subproject is concerned with the development of AI technologies for interactive mobile assistance robots to support everyday care.

Motivation: With 1.7 million employees, outpatient and inpatient care facilities have a share of almost 23 percent of all employees in the healthcare industry according to the BMWi in 2020. However, the increasing aging of the population is already having a major impact on care, as the need for care is expected to continue to rise in the future, accompanied by a sharp increase in the shortage of skilled nursing staff. Against this backdrop, robotic systems - as embodied AI - are seen as having enormous potential to contribute to relieving the burden of everyday nursing care by providing situation-adapted support services. Studies in nursing science show that the following fields of application for robotics in nursing are particularly significant, as nursing staff can be very effectively relieved here, leaving more time for important therapeutic and social contacts with the people being cared for: Transporting objects, access control, accompanying visitors and residents, assisting the cognitively impaired, or physically mobilizing the cared-for.

The aim of the subproject is therefore to provide particularly practice-relevant and easily demonstrable robotic support services for everyday care based on the extensive expertise of the TU Ilmenau in the robotics application fields of retail, home assistance and rehabilitation in the pilot project. The pilot project will therefore focus on the following use cases in assisted living facilities: Piloting visitors to residents and staff, accompanying residents on walks in the building, transporting commodities with the ability to autonomously use elevators and open doors.

Necessary prerequisite for demonstrable solutions is the development of basic AI-based robotic skills that can be replicated in other application fields of assistance robotics (e.g., retail or rehab): multisensory person perception and visual person recognition using Deep Learning, facial analysis and mask recognition (in pandemic times), and gesture recognition for user-robot interaction and person-centered robot navigation.

The focus of the developments will be on autonomous operation of elevators and opening of doors using a robot arm, which are basic requirements for use in unmodified real-world environments. A demonstration of the developed techniques by means of a mobile robot platform suitable for nursing homes equipped with the necessary sensors and manipulators in the premises of the cooperation partner AWO AJS gGmbH Regionalverbund Ilmenau is planned.