The school where robots learn to help people – and each other

In the future, there are many activities in which robots could assist humans – helping to feed people in care, assisting workers in factories with dangerous tasks, or assisting around the house with domestic chores.

Some of these tasks could be too dangerous for people to carry out, or there might be a shortage of trained workers.

Researchers at the newly-created Munich School of Robotics and Machine Intelligence, part of the Technical University of Munich, are developing tools to help robots and artificial intelligence (AI) help us in everyday life.

Robots in Vodafone Germany’s 5G lab. Photo credit: Vodafone

Founding Director Professor Sami Haddadin and his team want to make robots more flexible and smarter.

That requires AI – and also a fast mobile signal so that robots can share their knowledge with each other from hundreds of miles away, in real time.

To demonstrate how robots can share their knowledge via cellular networks, Professor Haddadin assembled 12 robots in the Munich School, 13 in the Pinakothek der Moderne (a modern art museum in central Munich) and a further 12 in Vodafone Germany’s 5G Lab in Dusseldorf, some 500 miles away.

The 5G Lab helps Vodafone and other companies to test devices using the latest technologies as soon as they are available, under realistic network conditions.

The robots were set a task: learn how to open a door with a key.

One robot at the Pinakothek was guided by hand and, after a few test runs and several minutes of machine learning, was then able to successfully carry out the required movements independently and accurately.

In an instant, all the other robots began to move like a robotic ballet.

They received the learning from their single apprentice across hundreds of kilometres and were able to use it immediately.

“This experiment is a milestone in the combination of robotics and AI. Networked robots can, for the first time, learn skills collectively – and faster than humans,” remarked Prof. Haddadin.

“The results from our research can be incorporated into applications with huge relevance for society and, together with our partners, we want to test these innovations in everyday life and put them into practice.”

Next for the Munich School is to develop robots and AI that can help senior citizens to maintain their independence for as long as possible, in a newly opened Centre for ‘Geriatronics’ – a concept involving the use of robotics and AI for the living arrangements and care of the elderly.