Seal of approval for robotics

Researchers at the Munich Institute of Robotics and Machine Intelligence (MIRMI) at the Technische Universität München (TUM) have developed an evaluation scheme that can be used to compare industrial robots and, in future, other systems such as mobile and humanoid robots or even robotic hands. Deputy MIRMI Director and Professor for the Perception of Intelligent Systems, Achim Lilienthal, is convinced: "The TUM MIRMI test methodology has the potential to establish itself as an industrial testing standard." This lays the foundations for establishing a seal of quality for the highly dynamic robotics market. "Knowing the performance of a robot system is a huge support for industry that wants to use robotic systems in the most targeted way possible." TUM MIRMI Executive Director Prof. Lorenzo Masia emphasizes: "Based on the original idea, I am sure that the AI Robot Safety & Performance Center of TUM MIRMI will develop into an independent national test center for robotics."

Start with one-armed robots

As a first step, the researchers examined and categorized single-arm robots from common manufacturers that are used in industry and research. Although many robotic arms look similar, they have their own individual strengths and weaknesses. The sensors, the motors and the actual brain of the intelligent machines, the control unit, differ. This results in systems that are very different in their basic capabilities. Some are powerful and precise in their movements, while others are gentle, flexible and sensitive in their interaction with their environment.

In order to make these differences and the continuous progress of robotics visible, researchers from the AI Robot Safety & Performance Center at TUM MIRMI have developed the so-called 'Tree of Robots'. Based on Charles Darwin's tree of life, it depicts the fundamentally different adaptation of different species to their 'habitat' - in this case not of living beings, but of robots. "To do this, we look at the basic abilities of a robot that are required in processes, such as how well it follows a given path, adopts a position, how gently it can make contact with surfaces and also how safe it is in dealing with collisions between robots and humans," explains Robin Kirschner, head of the laboratory.

For tactility alone, there are 25 measured values that describe how sensitive the robot is in physical contact with its environment. Among other things, the aim here is to find out whether the force that is supposed to be applied to a surface, for example, is actually higher than intended or whether a robot can cause injury to humans. From the pattern that emerges on a spider diagram based on the 25 measured values, even non-experts can see at a glance how sensitive the robot is.

Clarity about strengths and weaknesses

Depending on the performance of the individual systems, the researchers divide the robots into the classes 'industrial robots', 'cobots', 'soft robots' and 'tactile robots'. When it comes to a robot for surgery, the robotic arm is primarily required for precision, whereas for use in warehouses or production, strength and resilience are more important, i.e. the ability to perform certain movements many times in succession. "We combine existing movement metrics with our new tactile metrics and thus provide an overview of all the basic capabilities for physical interaction of a robotic system for the first time," explains Kirschner.