Adaptable Robots 'On Their Way' to the Home
Researchers have developed robots that learn to live with damaged parts in less than a minute, instead of the many hours needed bytraditional self-learning systems.
The system paves the way for robots to be used in awide variety of settings, coping with damage that occurs in the real world.
Possible applications include robots looking after theelderly, rescuing earthquake victims or doing housework.
Details are in the journal Nature.
We marvel at the robots we see in films: some try towipe out humanity, such as Ultron in the recent Avengers film, while otherslike C-3PO in Star Wars are helpful albeit slightly annoying. But they are thestill the stuff of science fiction, partly because if the slightest thing goeswrong with a real-life robot it usually stops working altogether.
In a step towards making the machines of the movies areality, French and US researchers have developed a learning algorithm thatenables robots to adapt very quickly when they are damaged.
Factory and laboratory robots have a strictlypre-defined way of operating. So if a component breaks, it normally has to bereplaced for the robot to continue with its task.
Most self-learning systems that seek alternative waysof continuing with the task are too slow, because they try out billions ofpossibilities. Using software that filters out the ineffective strategies, thenew system vastly speeds up the process.
The team demonstrated the principles with asix-legged, spider-like robot that found a new way of crawling across the floorafter one of its legs was broken. They also unveiled a robotic arm that couldadapt to a broken joint and learn a new way to drop a ball in a bin.
In both instances, the robots were able to learn a newstrategy to complete their task in less than a minute. Traditionalself-learning systems would have taken days.
According to Dr Jeff Clune of the University ofWyoming, the development represents an important first step toward robots thatare able to operate independently, outside of the carefully controlled confinesof a laboratory or factory floor.
"Having the kind of intelligent robots you see inthe movies is much closer than people realise. Our algorithm should inprinciple work on any kind of robot no matter how complex it is," he toldBBC News.
"If anyone can get a working C-3PO, our algorithmcould help it to deal with unforeseen situations and damage."
The new system uses a computer simulation to filterout all the possible solutions that will not work well. It then collects theones that are both effective and different from each other, so that the robotdoes not waste time testing out similar strategies.
Next, the robot tries out what the simulation predictswill be the best solution. If that fails, it tries out something entirelydifferent. It carries on doing this until it finds a strategy that works.
The research points the way to robots that are morerobust, adaptable and cheaper to maintain than today's machines, according tosenior author Jean-Baptiste Mouret, from Pierre and Marie Curie University inParis.
"If you have robots in your home they wouldprobably be expensive and you would not want them to stop working if a smallpart breaks," he told BBC News.
"Robots are used in factories because they are acontrolled environment and nothing unexpected can happen to them. We would liketo put these robots outside of factories to help in the outside world, whereanything can happen."
Alongside recent advances in artificial intelligence,such as the self-learningsystem developed by Google's DeepMind Technologies, this developmentcould see the emergence of new uses for robots.
"The things (DeepMind) is working on areamazing," says Dr Clune. "(Together) we will bring closer a futurewhere robots are helping humans.
"We should be sending robots into Fukushimainstead of asking human volunteers to take lethal doses of radiation. Robotsshould be putting out forest fires so we don't have to risk human lives andthey should be used to help us in our homes," he said.
Among the ideas for household robots are machines forcleaning, cooking,or loading and unloading dishwashers.
Most of the development of the system was done byAntoine Cully, a PhD student working with Dr Mouret. He says one of hisprincipal motivations was to help those who are infirm.
"I hope we can have robots that are assistantsfor the elderly," he said.
Dr Clune, though, believes this technology could haveapplications at the other end of the age spectrum.
He has very recently become a father, whose newbornchild has been keeping him awake into the early hours of the night. Heconfesses that he has spent the extra sleepless hours pondering "long andhard" whether he could use the new system to create a robot that cansoothe a crying baby.
"Necessity is the mother of invention, so maybethat will be my next invention," he joked.
But I was left with the impression that his commentswere not entirely in jest.
Follow Pallab on Twitter