When Denise Hebesberger hired Henry as a temporary worker at the Haus der Barmherzigkeit, a home for elderly and chronically ill people in Vienna, she worried about prejudice. Given the advanced age of many of the residents, ”we assumed maybe there [would be] some fears about interacting with him,” she says, but “observational data and interviews showed people are really quite curious and open-minded.”
Henry is a Scitos A5 mobile service robot produced by Metralabs, a German firm. The machine costs about €30,000, less than the median annual income for men in Austria.
“People were quite open as soon as they realized we don’t want to replace any staff members,” Hebesberger says. That was a big worry, she notes, because “there is this prejudice that there that there are many robots out there that are trying to take people’s jobs.”
Henry is part of the European Union’s STRANDS project, which aims to “produce intelligent mobile robots that are able to run for months in dynamic human environments.” The idea is to have fully autonomous robots capable of fulfilling some basic functions to support human workers.
Henry had two jobs. One was patrolling the care home. The other was “entertainment,” which involved hanging around at the entrance to the home and greeting visitors, who engaged with Henry through a touchscreen at chest level. “When he detected movement, he asked if they wanted to interact with him,” says Hebesberger. Many did.
The STRANDS project, which involves seven universities as well as G4S, a security firm, will run for another three years, with results from each trial informing the following year’s work. The consortium buys the robots off the shelf and then programs them with their specific purposes in mind. Henry, for instance, was given a louder, clearer voice, and larger text—both useful features in a care-home setting.
Bob of Birmingham
Around the same time that Henry was greeting visitors at the Austria care home, Bob started patrolling the corridors of the University of Birmingham. This robot worked under the watchful eye of Nick Hawes, who leads the STRANDS project.
“What we found is that at this early stage, it required more human expert input than we wanted. The software wasn’t working as well as we’d liked,” says Hawes. “But people in the office were pleased to have it around. They would say good morning to it and gave Bob a badge that showed it was a trainee security guard.”
Bob, like Henry, had a simple set of tasks, such as checking whether fire doors were open or closed, and whether important objects such as fire extinguishers were where they should be. Some 30 or 40 times a day, the robot took a break from these assignments and built 3D maps. That way, it could learn what had changed over time—a keyboard that had been moved, a monitor pushed askew. ”It’s going beyond the visual appearance,” says Hawes. “We call it the structure of the environment. That will allow it to detect when things don’t match the structure, [such as that] the top of the desk has been rearranged. The kind of things a human would notice as insecure. We want a robot to pick that up.”