At a recent car boot sale in the UK, Mark Tilden stumbled on
something he often looks for at second-hand markets – one of his famous
creations. Tilden, formerly a roboticist the Los Alamos National
Laboratory, is a toy designer and he was now holding in his hands one of
the humanoid bots he invented for children, “Robosapien”. Delighted
with his find, he bought it for, “something like five quid”. The bot
would give him the perfect opportunity to see exactly how it had been
used.
“This toy had been played with to death. I was amazed,” he
says. “It was filled with sand, it was filled with Plasticine, it had
make-up still on it, rusted batteries – it had obviously been taken into
the bath-tub – and that was fascinating because all of a sudden you
realise that someone had loved this toy to the absolute extent.” Sure,
it had been up for sale, but this bot had a good life.
What was it
about this robot that had appealed so much to its owner? It’s a
question that Tilden and other roboticists think is important – not just
for toy design, but the future of robotics. For too long, robots have
suffered from an image problem. They are often perceived as mechanical, cold and threatening in our culture
and it’s difficult to reverse that impression. But this view of robots
could be changed if they were designed to appeal to us with the same
familiarity and, indeed, personality that our childhood toys once
possessed.
Children’s affection for Robosapien could be
explained by the bot’s ability to display these characteristics, argues
Tilden. His robot was never designed to seem super smart or unreasonably
clever, but to have foibles and quirks that would entertain children
and engage their imaginations. For instance, the toy’s 67 pre-programmed functions
include belching, rapping and dancing. Seeing the Robosapien as a pal
was far more important than seeing it as a hyperintelligent, futuristic
machine.
Could other successful toys provide similar cues for
robot designers? Perhaps – and it needn’t even be toy robots. Take
Cabbage Patch Kids, for example. A highly successful line of dolls,
Cabbage Patch Kids appealed to children’s emotional intuition through
their insistence on being taken seriously as infants that required love
and attention. They needed their nappies changing, as one advertisement explained,
and came with documentation like birth certificates and adoption papers
which positioned them within a mock bureaucratic world of parenting. As entertainment scholars have noted,
it was the emotional responsibilities of owning a Cabbage Patch doll
that made them persistent as toys – kids couldn’t put them down because
they had a duty to look after them, a duty which they understood
instinctively.
How do we get from here to robots? A few other
toys will help to explain. In the 1990s, the electronic sophistication
of toys was burgeoning and offered new opportunities to exploit
children’s empathetic instincts. Tamagotchis were small egg-shaped mini
computers with a square LCD screen displaying an animated pet.
The
onus was on the child to look after this pet by feeding and playing
with it. More than 76 million Tamagotchis have been sold worldwide since
their arrival in 1996. The phrase “Tamagotchi effect” was coined to describe the strong emotional attachment to virtual agents exhibited by Tamagotchi owners.
Newspaper reports at the time marvelled
at how children expressed extreme outpourings of grief when their
“cyberpet” finally succumbed to that great leveller of all, death.
Tamagotchis required children to carry out a duty of care, like Cabbage
Patch Kids, but the difference was that the consequences of bad
parenting would actually be played out.
The Furby, launched in the
1990s, was designed to evoke similar emotions. Furry, talkative and
exhaustingly needy, Furbies yearned for love and care. In this Radiolab podcast,
Furby co-creator Caleb Chung explains that Furbies were designed to
appeal to human beings’ innate sense of compassion by sounding scared
when held upside down, or by quivering at loud noises.
Quaint, you might think, and something that only small children would do. But consider the soldiers in Iraq and Afghanistan who gave funerals to combat robots when they were irreparably damaged. That machines provoke strong emotional connections with us is not the preserve of children.
All of this is now directly informing the work of
robot designers around the world. Aldebaran Robotics, headquartered in
Paris, has learned a lot from the world of toys.
“When we saw that
children were able to have a very strong connection with a Tamagotchi, a
very simple device, we knew that it would be possible to create a much
stronger connection with a robot that had a humanoid shape, an
expressive voice and expressive gestures,” explains Aldebaran’s research
director, Rodolphe Gelin. “Toys demonstrated to us that this was
possible.”
The company’s most famous bot is Nao – a humanoid robot used in educational settings.
Nao is intentionally the size of a teddy bear, says Gelin, so as not to
be too large and “invasive”. Nao also has LEDs around its eyes and on
the sides of its head that animate to provide further cues, such as
blinking, which suggest it is alive. Even simple things, like making
sure Nao didn’t have a lot of visible metallic parts, helped to
naturalise it.
Aldebaran has also crammed the robot full of
sensing capabilities which mean its behaviour can subtly adapt to the
needs of human companions. Cameras and image analysis can distinguish
your gender, or mood, for example, and 3D scanners read body language.
Studies
that explore how children interact with robots have shown that this
level of social intelligence is crucial for supporting long-term
relationships with machines. Lola Canamero, a researcher in adaptive
intelligent behaviour at the University of Hertfordshire, UK, agrees
that children are generally very willing to initially suspend their
disbelief when greeting a robot. However, that suspension of disbelief
may not last very long unless the variety of interactions can continue
to engage the child.
“If children see that the robot is actually
responding to what they do and not just performing random actions, that
keeps their interest for much longer and persuades them to continue
interacting with the robot,” she says.
The results of all these
efforts to improve the social capacities of robots speak for themselves.
For instance, Gelin recalls visiting a nursing home in France to ask if
the residents there, who had never seen Nao, would be interested in a
robot companion. The elderly people scoffed at the idea. Gelin told them
he of course understood – but before he left, he pulled Nao out of his
bag. The mood in the room transformed. Suddenly the nursing home
residents were fascinated and intrigued. “Can he sing? Can he talk to
me?” they asked.
It’s this eagerness for companionship that Gelin says robot manufacturers must now appeal.
“What
is most important for us is to have a robot which understands your
emotion and which can itself express emotion,” explains Gelin. “What one
generally imagines for a robot, that he will perform tasks, clean the
house, bring you a beer – for us these things will come later. The first
thing is to have robots which are accepted at home.”
After all,
when we were children, we had a special connection to our toys. They had
personalities, could follow us on adventures and proved to be the
ultimate companions. Perhaps the most promising robots of the future, then, will just be toys all grown up.
If you would like to comment on this article, or anything else you have seen on Future, head over to our Facebook or Google+ page, or message us on Twitter.