Hiccups, bugs and public failures are an inevitable part of the deployment of any tool in the real world, but robots must also be designed to account for sometimes unpredictable human interactions.
"Social robots, if they're engaged in a public sense — even in a limited public sense — the design has to include considerations for social interactions," said David Harris Smith, associate professor at McMaster University's Department of Communication Studies and Multimedia in Canada.
Smith knows this firsthand. Together with associate professor Frauke Zeller of Ryerson University in Canada — and a cadre of other scientists, artists and engineers — he developed HitchBOT, a robot designed to travel across a country by "hitching" a ride from friendly humans. Completely immobile, HitchBOT was designed with an LED "face," a hitchhiking thumb and the ability to respond to simple voice questions. Creator Frauke Zeller said she wanted to create the impression that HitchBOT "is a helpless robot and challenge people to become active and engaged."
The HitchBOT experiment came to an end after the bot was found dismembered and destroyed in a Philadelphia alley. With the dawn of the everyday robot age, destruction is a necessary part of creation.
"In terms of designing these robots, we have to take a step back and have people decide," said Zeller. The key question is how — or if — we actually want to live with robots in our midst.
"We're really dealing with some deep-seated cognitive processes when we start mixing humans and robots," Smith said. Humans empathize with less-capable creatures and conversely fear hyper-capable creatures, epitomized by pop-culture robots like Terminators and even Transformers. Designing friendlier robots requires an element of what is called participatory design.
"If it's gonna be in the context of engaging with human co-workers ... there needs to be some kind of co-design in their spec, and that affects how it will fit in," Smith said.