In a side-street just off Broad Lane, opposite St George’s churchyard, there’s a building you’ve probably wandered past a hundred times. With its bullish, haughty Victorian frontage, the Sir Frederick Mappin building looks faintly medical; a training college? A natural history museum, maybe? In actual fact, it’s an incubator. Deep inside, the next generation of smart robots, designed to be our new BFFs, are waiting to be born.
For the last three years, Professor Tony Prescott, Director of the Sheffield Centre for Robots and Professor of Cognitive Neuroscience at the University of Sheffield, has been part of a group of academics and designers taking new theoretical and practical advances in robotics and artificial intelligence into the average home. The group’s first prototype biomimetic robot (that is, a robot which works by mimicking a system from nature), MiRO, has just been released, and is a very different kind of help to the average automatic vacuum-cleaner.
MiRO is an assistive robot which won’t perform menial tasks but would, Tony says, ‘provide the social benefits of a pet animal’; it’s an emotional and social crutch which could help to improve the quality of life of people dealing with loneliness and social isolation but who couldn’t cope with a real puppy. To build MiRO, the scientists have had to research some profound questions: What is friendship? How do you synthesise it in a lab? And once you have, what do you do with it?
If you’re worrying that we’re wandering slowly toward the plot of The Earth Dies Screaming (which, if you’ve not seen it, is as cheery as its title suggests), you should know that MiRO is about 1ft tall, and looks part-dog, part-Japanese crime-fighting cartoon sidekick. If it does bring about the end of human civilisation, at least it’ll be the most adorable Armageddon possible. MiRO wouldn’t be a pain in the neck like renowned pod bay door-refusenik HAL from 2001: A Space Odyssey; not only would MiRO open the pod bay doors, he’d probably make you a packed lunch for your big day in space.
So how did they go about creating it? After sorting out MiRO’s basic senses – it can see, hear, feel and navigate around without launching itself down a staircase by accident – the team had to look at why some people are so mad about their pets that they end up treating them as particularly hairy surrogate children. Or, as Tony puts it with the specificity of the seasoned academic, why they have ‘long-term interactions with companion animals’.
‘One is that there’s some shared experience and shared memories, so you certainly feel that the animal you spend time with knows who you are and remembers some of the things you’ve done together, and forms a bond,’ Tony says. ‘Most robots right now, they don’t know there’s anything else out there; they do what they’re going to do, but they’re not aware of people, or even very much of other things. That’s a big step for artificial intelligence: to create that kind of social intelligence.’
For MiRO, that starts with modelling a dog’s perception of the world and its relationship with it. ‘There would be notions of you being the leader; as a dog it would be looking to you to know what it should do next. When we try to build this into a robot, the first task is to get the robot to even recognise that there’s a person. That’s non-trivial,’ says Tony. ‘We’re going to be able to put those onto robots so that the robot can recognise, hopefully, who you are, maybe tell from your voice or the way you look whether you’re happy or sad and maybe do something appropriate in relation to your mood.’
When it’s put like that, it feels like this tiny plastic dog might grow up to be something like Metal Mickey inhabited by the spirit of Terminator’s Skynet, but be reassured that MiRO’s completely benign. Finding an aesthetic which communicated MiRO’s purpose without freaking anyone out was the job of Sebastian Conran, award-winning designer of satisfyingly utilitarian homeware. Tony says that most people have had ‘a very positive reaction’ to MiRO, which is no mean feat in itself. It could have fallen into a place all roboticists fear: the uncanny valley.
Imagine a graph which plots positivity of human response to a robot (running from, say, a confused shrug of indifference to proposals of marriage) on one axis against the lifelike qualities of the robot itself on the other. At the bottom end, you have the robotic arms used on assembly lines (which, you’ll notice, nobody has ever attempted to cast as the hero of a whimsical children’s film; they aren’t particularly cutesy) all the way up to a healthy human being. As the design and function of a robot becomes more lifelike, the more a human can identify with it, bond with it and project emotions onto it. Until, that is, you hit the uncanny valley. There, the trend reverses sharply; the robot is lifelike enough for us to expect lifelike behaviour from it, but sufficiently un-lifelike to be viscerally upsetting in the same way as ventriloquists’ dummies, zombies and corpses can be.
‘The reason for that is that we’re very sensitive, particularly to cues from other people,’ Tony says. ‘If someone’s acting a bit strangely, or looks a bit odd, then we might think that maybe this person is acting strangely for a reason. Maybe they have bad motives.’ This deep neurosis in the human psyche is a problem which robotics engineers and thinkers (as well as sci-fi writers) have been grappling with for decades, and which Tony and his colleagues found a way around by pursuing a consciously cartoonish animal aesthetic: ‘We are finely tuned to pick up those signals from people; I think we’re less finely tuned for animals. But I think there’s still a bit of an issue, and I don’t want to make robots that are just like animals to look at but move a bit differently, because I think that could potentially be creepy.’
That’s where Sebastian’s expertise in designing appliances was particularly valuable. ‘People instinctively approach our MiRO robot and like to pat it and touch it, which is all the things it’s designed to do,’ says Tony, ‘but it’s also very obviously a robot. Okay, it looks a bit like a puppy, but I’m not going to expect it to chase sticks.’
Robotics is moving at such a pace these days that Tony feels we’re ‘at a tipping point’ with social intelligence tech, and Consequential Robotics hope to have a retail version of MiRO on the market in the next year or so: ‘The idea is that people will spend maybe up to £1,000 to buy a pet animal, so for a similar amount you might buy a pet robot.’
Imagine: a world in which you’ve always got someone to watch Ru Paul’s Drag Race with, someone to moan about work to, someone to trundle through life alongside; and all it asks in return is a warm plug socket and an occasional once-over with a J-cloth. The future’s here, and it’s way cuter than expected.