Hospitals are, even at the best of times, a sort of ordered chaos. Patients, overwhelmed by it all, wait or wander looking for care. Nurses and doctors, in the thick of it, juggle deep knowledge with pressing need and dozens of immediate tasks. It’s a lot to manage, which is why researchers at MIT’s Computer Science and Artificial Intelligence Laboratory trained a small NAO robot to evaluate the situation and offer suggestions when asked.
The robot analyzed movements of people in the labor ward, and then came up with good and bad recommendations for what to do next. The bad recommendations served as a check to make sure the robot wasn’t just making arbitrary suggestions, as well as a test to make sure nurses weren’t blindly accepting the robot’s advice. When asked specifically for a good decision, it would offer one that hospital workers would accept 90 percent of the time, and when asked specifically for a bad decision, it would offer one the hospital workers rejected at a similar rate.
Scheduling can be tricky for lots of people in lots of circumstances. By adding in a robot that can analyze scheduling needs, hospitals could make better informed decisions. At least, so long as the scheduling robot only augments human workers, and doesn’t replace them.
Watch a short video about it below: