Technology is increasingly a part of our homes and daily lives, but as these advances become more sophisticated, with researchers now trying to mimic the workings of the brain, could some future super-robot become a threat to humans, asks JOHN HOLDEN
THE HUMAN response to robots has always been ambiguous: science fiction movies portray an even blend of good and evil machines. If they’re not trying to destroy the human race like Arnie in Terminator, they’re showing us the better part of ourselves, like Pixar’s sensitive droid in Wall-E (above). Utopian or dystopian, robots draw people into the cinema in their droves.
The thinking robot is, conceptually, the ultimate labour-saving device and anything that reduces our workload is appealing. Still, there is always the notion that machines shouldn’t develop too much intelligence for fear they might take over altogether.
What most of us don’t realise, however, is that humans and machines are moving closer together all the time.
“The fact is people react socially to computers whether they want to or not and there is research that shows this,” explains Thomas Holz of the UCD Robotics group.
A research project carried out at UCD placed a number of people in front of computers that acted as trainers. The computers delivered instruction on a subject that was completely new to the volunteers. After the training session a second computer was used to test the subjects on what had been learned.
Next the participants were divided into two groups – half using the computers that had provided the initial training and half with a third wholly new computer. Finally, both groups were asked to rate the success of the training programme.
“The ones interacting with the computer that had taught them were less likely to blame the computer’s teaching skills on how well they did in the test,” says Holz.
“Essentially they were being polite, stating that if they did badly it was their own fault. They humanised the machines. It’s a subconscious thing but we all do it sometimes.”
Other research studies into how we intereact with robots have shown that a robot’s appearance will affect human trust. “People will trust robots more with a baby’s features,” says Holz.
“If you design a robotic medical consultant, for example, you want your users to trust that robot. So you give it what we perceive to be more trustworthy human characteristics – big eyes, big forehead and a small chin.”
While we all subconsciously socialise machines, some Irish roboticists are consciously imitating the human condition to create more advanced robotic devices. In NUI Galway the Bio-Inspired Electronics and Reconfigurable Computing (Birc) research group have artificially mimicked the neurological structure of the human brain.
“In conjunction with the Intelligent Systems Research Centre at the University of Ulster, Magee, we have developed an integrated circuit called a hardware spiking neural network, which can be used for controlling autonomous robots,” explains Dr Fearghal Moran of Birc.
“It is modelled on the human brain but in very simple form. Our brains have millions of interconnected neurons. These neural networks would comprise of only about 50. But the network can be trained to perform a particular function using a set of training data.
One of the goals of the research is the control of small, autonomous robots which can operate in remote, unsupervised environments, like in space and remote search and rescue.”
As this technology is improved and networks are built with ever-greater numbers of neuron models – closer to the human brain – are achieved, what kind of super robot could be created?
“A machine with a very complex view of the world and a very large quantum of autonomous skills,” says Moran. “But without self awareness it would still be no danger to humanity.”
Such potential risks – involving a robot effectively going human – are of concern to roboticists. “Ethics is an issue,” says Dr Gerard Lacey of the School of Computer Science and Statistics in Trinity College Dublin.
“All of us who work in robotics want to build systems that are intrinsically safe. It may be a little early though for true robotics ethics, as systems would have to develop self-awareness and we are still a long way from that,” he says.
“Robots and people are sharing the same space and co-operating to solve human problems,” he adds.
“Many of the technologies designed to assist humanity are just that: assisting. People don’t want to feel controlled. We do user studies on these kinds of issues all the time.”
Current robotics research has delivered many positive outputs, especially in enhancing the lives of those with physical disabilities or other debilitating conditions. In the Magee campus of the University of Ulster in Derry, true human/robot co-operation is being realised.
“One of our main projects is the development of technology to effectively drive performance and enhance the ability to control a wheelchair using brain signals alone,” explains Professor Martin McGinnity, director of the Intelligent Systems Research Centre at Magee.
“This technology is designed for people with conditions such as ‘locked in’ syndrome. Extracting ‘intention’ from the brain is a difficult challenge. But we’ve developed algorithms to do this.”
Cyber homes
A EUROPEAN TEAM of robotics experts led by UCD researchers are working on new technology that could turn the home into something from science fiction.
Imagine a dishwasher that knows when it's full, so it switches itself on. Or a vacuum cleaner that waits until everyone is out before it gets to work.
The €2.5 million Rubicon (Robotic Ubiquitous Cognitive Network) project, is the first EU-funded robotics project to be co-ordinated by an Irish university. The team aim to bring cutting-edge technology into the home by creating intelligent robotics networks that co-operate with each other.
"This is a new type of computer system – a robot ecology of robotic devices embedded in everyday environments, that co-operate in the perform- ance of complex tasks," says Dr Mauro Dragone of the UCD Clarity Centre for Remote Sensor Web Technologies.
A home could have a central brain system that connected all household appliances together. The appliances could co- operate and learn from each other.
Dr Mauro Dragone was initially inspired by the movie Avatar. "In the film, the blue creatures attach their heads to tree roots on their home planet," he says. "Then they share new knowledge with the brain of their planet. I figured we could do the same thing with robots. We are working on building a system of robots that are connected in a domestic setting so that the system can learn how to help the human without the human having to instruct the system," he says.
"Enabling robots and sensors to seamlessly co-operate in such ecologies is an important challenge for robotics RD, with applications in ambient-assisted living, environmental monitoring, and security," he argues.
"Healthcare is a particular area of emphasis for the project, but we hope that our work will see even more practical applications in the home."
John Holden
See fp7rubicon.eu