Joint project RoSylerNT

  Copyright: © IRT

Learning robot-assisted systems for neuromuscular training

01/09/2017

Key Info

Basic Information

Duration:
01.09.2017 to 31.08.2020
Acronym:
RoSylerNT
Group:
Biomedical Systems
Funding:
BMBF

Contact

Name

Maike Stemmler

Head of Biomedical Systems

Phone

work
+49 241 80-27508

Email

E-Mail
 

Motivation

Logo RoSylerNT

Regular physical training offers great potential in the prevention and treatment of age-related and chronic diseases. Additionally, physical activity can positively impact subjective well-being. However, effective physical training requires high training intensity and thus high muscle forces, which can lead to uncontrolled overloads and damage to the musculoskeletal system. To avoid these non-physiological loads, force control through appropriate motion guidance is essential. Assistive systems that apply optimal training stimuli and minimize the risk of overloads can provide valuable support in this regard.

 

Project Goals and Methods

Sponsored by the Federal Ministry of Education and Research

Within this project, robotic systems should be empowered to actively and safely interact with humans, applying significant forces, for example, during neuromuscular training. To achieve this, it is necessary to ensure that the robot responds appropriately based on the current state of the patient and the environment. However, individuals' responses to stimulation or orthopedic interventions are highly individualized. Therefore, it is necessary to capture the posture, movement, and load of the person during training. Based on this information, model-based and model-free, self-learning control strategies is developed, with the human as the process to be controlled taking center stage, and the robot taking on the role of actuation. By utilizing a shared knowledge base, the robot learns the optimal control inputs to achieve the application-specific goal.

 

Innovations and Perspectives

The fundamental skills developed within the project open up a wide range of new application scenarios outside the traditional industrial sector. By using self-learning control algorithms with the explicit consideration of boundary and side conditions, the movement ability and physical capacity of older individuals or individuals with physical limitations can be taken into account, ensuring adequate support.

 
Project partner