Researchers at the Munich Institute of Robotics and Machine Intelligence (MIRMI) at the Technical University of Munich (TUM) have introduced a classification system intended to evaluate the sensitivity and physical interaction capabilities of robotic systems. The initiative aims to establish a standardized method for assessing robots’ suitability for tasks involving direct contact with humans and their environment.
The system, developed by MIRMI’s AI Robot Safety & Performance Centre, begins with the evaluation of single-armed industrial robots. These systems, although similar in appearance, differ considerably in sensor technology, motor control, and processing capabilities. The researchers assessed a range of performance indicators, including motion precision, positional accuracy, contact sensitivity, and collision safety.
The evaluation framework includes 25 distinct measurements specifically focused on tactility—how a robot manages physical interaction. These measurements examine, for example, whether a robot applies appropriate force during contact and whether it can safely operate in proximity to humans. The resulting data is visualized using a spider diagram, offering a comparative profile of each robot’s sensitivity.
To contextualize these differences, the researchers created the “Tree of Robots,” a taxonomy inspired by Darwin’s Tree of Life. This conceptual model maps various types of robots—such as industrial robots, collaborative robots (cobots), soft robots, and tactile robots—according to their functional adaptations to different application environments.
The classification system combines existing motion metrics with newly developed tactile metrics to provide a comprehensive overview of a robot’s capacity for physical interaction. The long-term goal, according to MIRMI leadership, is to establish an independent national testing center and develop an industry-wide certification standard for robotic systems.
The research is part of MIRMI’s broader work in robotics, perception, and artificial intelligence, involving over 70 professors at TUM. The institute’s efforts are directed toward advancing intelligent systems in fields including healthcare, manufacturing, environmental monitoring, and mobility.
Photo credit Andreas Heddergott / TU Muenchen
