The collaboration between the DLR’s Institute for Robotics and Mechanics, where Patrick van Smagt is conducting research, and the Department of Neuroscience, the workplace of John P. Donoghue, began several years ago and has led to the development of a globally unique support system that allows paraplegics to control a robotic arm using only their thoughts. To build the system, the researchers developed learning-enabled software that translates signals from the patient’s brain into control commands for the arm.

In 2011 a female patient who had been paralysed from the neck down for 15 years managed to use the arm to insert a straw into her mouth and drink by herself for the first time since her stroke. Extensive training was unnecessary. The patient only had to imagine moving her own arms in the same way, which generated signals in the motor cortex of her brain. A small implant in her skull, co-developed with Brown University, transmitted these signals to a computer, where they were transformed into the desired control commands by a learning algorithm that the researchers have continuously optimised.

“During these experiments it is of course crucial that the robot does not pose a danger to the test person,” says Van der Smagt. Safety is ensured by sensors on the robot’s arm, which constantly check whether the arm makes unwanted contact with the environment. If this occurs, a special programme immediately intervenes, causing the robotic arm to go slack within milliseconds. This technological breakthrough attracted international attention and has great potential to simplify the lives of people with disabilities.