MeRoSy

 

Human-Robot Synergy: Learning and Adaption in Human-Robot Evolution

 

Synopsis

The MeRoSy project focuses on the research, development and implementation of machine learning and adaptation methods for cooperating robots and assistive robots based on the measurement of human head movements and responses.

  Copyright: © EVADO Understanding head gestures for robot control  

Challenges

Currently, robots are mostly controlled via handheld operating devices, which makes it impractical to simultaneously perform a manual activity. People with physical limitations, such as the paralysis of their upper extremities, are not able to use such an operating interface.

Goals

The goal of the MeRoSy project is to research and implement alternative input options for controlling robots via head movements. The desired assistance system relies on machine learning methods with the goal of solving new tasks or learning and adapting existing solutions to new boundary conditions. To achieve this, the problem-solving behavior of humans was analyzed on the basis of event sequences and replicated. Based on existing solutions and by evaluating human interventions, this should lead to an evolutionary extension of the robot's capabilities.

Methodology

One focus of the IAW's work is on the intuitive design of human-robot interaction, which ensures conformity between human expectations and the robot's behavior. Based on the TAM 3 model, a technology acceptance model for human-robot collaboration was established. Furthermore, the ethical, legal and social implications of human-robot collaboration were discussed in expert workshops and subsequently analyzed.
Furthermore, the ergonomic design of human-robot interaction with the aim of reducing mental and physical strain is supported. In this context, studies were conducted based on the stress-strain concept according to Rohmert (1975) to assess the physical stress (motion tracking) and the resulting physical strain (surface electromyography).

Funding

The research project MeRoSy (funding code: 16SV7190) was funded by the BMBF – Bundesministerium für Bildung und Forschung (Federal Ministry of Education and Research). On behalf of the BMBF, the VDI/VDE-IT took over the project management.

Project Partners

University of Bremen, Institute for Automation Technology IAT

Westfälische Hochschule, Sensortechnik und Aktorik in der Medizintechnik, Fachbereich Elektrotechnik und Angewandte Naturwissenschaften

pi4 robotics GmbH, Berlin

Vorwig GmbH, Bremen