“The intelligent robots are coming,” “Man thinks – the robot controls,” or “AI will change our lives?” – Theses like these are in the media. KUKA used the European Robotics Week 2018 to discuss the “Future of Robotics” with renowned experts in a media seminar.
The German government alone plans to invest three billion euros in artificial intelligence (AI) by 2025 – a term that no one really knows exactly what it means. At least that’s what Torsten Kröger says. He is professor at the Karlsruher Institut für Technologie and is considered one of the most renowned robotics experts in the field of artificial intelligence:
“The term artificial intelligence is not clearly defined in either science or literature.” Rather, it is individual buzzwords that are driving the discussion around Industrie 4.0. “The reason for the current hype is one area of AI: machine learning. This technology creates added value”. Language assistants or digital image recognition, for example, work with this technology. Compared to humans, robots can perform such tasks relatively well.
Who can do what?
But this is also one of the few advantages the robot has over humans, says Kröger. Although industrial robots are powerful, precise and repeatable over thousands of working hours: “In terms of motor skills we are still unable to do anything in robotics. An algorithm can only execute one thing at a time, neural networks are not flexible. They also only work if you have a collection of a lot of data with which the networks were previously trained.”
Stefano Stramigioli, head of the Institute of Robotics and Mechatronics at the University of Twente, also knows the advantages of humans over robots: “As long as I live, no robot will come to my home as a plumber. The human hand alone is much finer and can work much better than a robot”. The advantage of humans also lies in their efficient use of energy. “When a human wants to work precisely, he tenses his muscles and when he doesn’t need them, he relaxes them again,” says Stramigioli. He sees the greatest challenge in making robots smaller and more energy-efficient in the future.
The versatility of the human being
Kirsten Albracht, a professor at the Deutsche Sporthochschule Köln, is also involved with human musculature.” The human being has a very complex structure and can act in many ways. They can open doors or climb stairs. Integrating this versatility into a robot becomes one of the greatest challenges. The design of the human body is extremely clever.”
Philippe Lorenz works for the Stiftung Neue Verantwortung in Berlin and heads the project “Artificial Intelligence and Foreign Policy”. He adds another important point to his versatility: “Robots still lack the ability to draw causal conclusions completely.” In humans, on the other hand, even small children can recognize the connections between cause and effect.
Conclusion of the discussion: Until now, robots have been repetitive, working with the same precision and repeatability. The requirements of the future are different. If robots are to be used in other areas, they must become more flexible. Machine learning can help here.