Optimizing Autonomous Control of Robotic Systems that Interact with Humans
In Part 2 of Machine Design’s interview series with Dr. Brokoslaw Laschowski, a research scientist and principal investigator at the Toronto Rehabilitation Institute, describes how his lab is developing control systems for robotic prosthetic legs along a spectrum of autonomy, ranging from fully-autonomous control using machine learning to neural control using brain-machine interfaces. Simply put, he specializes in giving robots intelligence, powered by either artificial brains or human brains.
Among Laschowski’s long-term goals is an ambition to conduct the first clinical trials to assess what level of control individual patients prefer. Some patients may not be instinctively inclined to hand over control of their movements to robots powered by artificial intelligence, Laschowski said. Rather than cede control, some patients may prefer to maintain control over the robot. One way to achieve this is by using a neural interface, either at the muscle or brain level, so that the patient can decide how the robot should behave simply by thinking.
Comparison With Autonomous Driving
One way to conceptualize his research, according to Laschowski, is to think of the various levels of driving autonomy. Most people have experienced human control when driving, Laschowski explained. In this case, the driver is handling the steering wheel. The opposite end of the spectrum are fully autonomous vehicles, where the car would drive itself using artificial intelligence without any human intervention. He noted that some people may think they like the idea of fully autonomous cars, but since such vehicles don’t yet exist, they cannot know for sure.
READ MORE: Transforming Prosthetics: The Nexus of Technology and Athletic Performance
Whether driving an electric car or walking with a robotic leg, there exists this spectrum of autonomy, ranging from fully manual to fully autonomous. Different people will likely prefer different levels of autonomy. “The automotive industry is studying this and seeing what level of autonomy individual persons prefer,” said Laschowski. “We’re doing the same thing with robotic legs.”
The long list of potential users of his robotic technologies, including patients with amputation, spinal cord injury, stroke, multiple sclerosis, Parkinson’s disease and many others, opens a wealth of opportunities within the field of rehabilitation medicine. However, in addition to assisting clinical populations, Laschowski said his robots could also be used to augment “healthy” users such as workers in firefighting, search and rescue, military, and manufacturing and give them superhuman performance.
READ MORE: Prioritizing Human Safety in Industrial Robot Design
Watch additional parts of this interview series with Dr. Brokoslaw Laschowski:
