You have to assume that Dr. Goldie Nejat has never seen the Terminator movies nor was a fan of Battlestar Galactica, any of the other fictions depicting a dystopian future after the likes of a robot armageddon. Otherwise, why would she construct a human-like robot that can convey and decipher emotion?
And then name it Brian.
Over-used science fiction concepts aside, Nejat and her team of grad students at the University of Toronto’s autonomous systems and biomechatronics lab has only positive intentions for Brian. The robot is designed to assist the elderly – not to physically do tasks for them, but to prompt them to remember important tasks and to maintain a healthy daily routine. More like a motivational coach than a futuristic butler.
The lab is working with Baycrest, a Toronto-based elderly care and research facility. Residents and staff there are helping take part in a pilot project that gauges how the aged will react to a robot that prompts them with friendly reminders. Brian isn’t ready to autonomously travel around a floor yet, but the hope is he (it) will visit various patients and tell them when it’s time for dinner, or encourage them to exercise.
Related Story: Gadgets for grandpa – tech tools to ease the ageing process
Nejat was motivated to build Brian by Canada’s ageing population. The Baby Boomer generation begins retiring next year, and that huge demographic is expected to tax the health care system as it continues to age.
Brian could be a hint of things to come. Technologies providing robotic care-givers and other assisted-living methods could soon be a booming industry.
“We wanted to take care of those large numbers of Baby Boomers,” Nejat says.
Baycrest is always exploring new technology that could help the elderly live fuller lives, says Bianca Stern, the director of culture and heritage. While those close to 100-years-old may be opposed to the idea of an assistive robot, those between 70 and 80 are more open to the idea. As Baby Boomers age, they will be even more receptive to robots thanks to a lifetime of exposure to technology.
Acceptance will also vary by culture.
“There are all these moral and ethical implications for social connectedness and human beings versus non-humans,” she says. “How cultures will adopt this is a whole other question.”
Residents at Baycrest haven’t met Brian first-hand yet. But they have watched some videos of the robot in action as part of a pilot project to see how they might react to it. Eventually, Brian will be brought in to meet some volunteers – first, the robot needs to be attached to a lighter and more mobile base.
Currently Dr. Nejat’s team is working on Brian’s brain. They’re upgrading his emotional detection software to determine mood through tone of voice, body language, and facial recognition methods. It’s a work in progress.
Also, Brian is learning how to react appropriately to a person’s mood.
“If you’re frustrated, the robot will try to provide a happier emotion and more excitement on its part to minimize your stress level,” Nejat says. “It provides encouragement; it says sentences like ‘I know you can do this.’”
Brian is equipped with a speech synthesizer and a computer generated voice. Emulating a deeper man’s voice, it expresses emotion by varying pitch and tone.
The robot’s latex face is human-like, but it wouldn’t fool anyone. Servo mechanisms move this mask into different positions to look happy with a smile, sad with a frowning gape, or stern with a furrowed brow.
“We want the person when they first see the robot to know it’s a robot and not a human,” Nejat says.
The robot also has the capacity to learn in specific contexts. For example, it could assist a senior in playing a memory-testing card game by remembering the location of a certain card. Then it could provide hints to the senior on how to succeed.
Brian cost about $20,000 to build, Stern says. That price will come down if more robots like Brian are constructed and the process becomes more efficient. If the robot results in higher quality care for patients, then it’s worth having in the director’s eyes. A cost-benefits analysis hasn’t been done yet because the robot is still in prototype phase.
The emotional detection and response software could have other applications beyond health care. Potentials users include the gaming industry – imagine a game that could adjust its difficulty based on how calm you appeared while playing. One other area Dr. Nejat is considering is search and rescue.
“It could also be a robot used in human-robot teams,” she says. “Rescue robots could work alongside rescue workers for example, after a major disaster.”
That disaster won’t be caused by any robot army constructed at her labs, Nejat assures.