Page content

Ulster Scientists Explore Intelligent Robot Healthcare

28 October 2010

Scientists at the University of Ulster’s Intelligent Systems Research Centre (IRSC) are involved in a major European research project, which could pave the way for the development of the next generation of intelligent robots.

Roboticists at the Magee-based research unit are part of an international consortium of leading neuroscientists, psychologists and machine learning researchers involved in the £6 million interdisciplinary IM-CleVeR project – an acronym for Intrinsically Motivated Cumulative Learning Versatile Robots. The overall objective of the project is to develop real time intelligent robots that learn from their surroundings.

The Principal Investigator on the project, Professor Martin McGinnity, Director of the IRSC, says the project is of huge technological and scientific significance with potentially far reaching implications.

“Robots can already be programmed to perform repetitive tasks. The IM-CLeVeR project  aims to develop a new methodology for designing real time intelligent robots that can be intrinsically motivated to learn new skills and then to reuse these skills to carry out increasingly complex tasks. This could have many applications, for example providing assistance to patients with Parkinson’s disease or helping elderly people who live alone.”

Dr Joan Condell, a Co-Investigator, explains that the central idea of the project is that instead of directly programming, training or evolving a set of specific skills in robots, the robots should be given developmental programs that allow the autonomous development of skills on the basis of prolonged periods of interaction with the environment under the guidance of intrinsic motivations. Robots could then use this acquired general ability as a building block for the solution of specific tasks relevant to the robot's users.

She says that during skill-acquisition, the robots will behave like children at play.

“Children at play carry out activities driven only by intrinsic motivations such as curiosity. These activities allow them to acquire knowledge and skills autonomously which they can use later in life. During skill-exploitation, the robots will exhibit fast learning capabilities and a high versatility in solving tasks defined by external users due to their capacity of flexibly re-using, composing and re-adapting previously acquired skills.”

The basis of the IM CleVeR project is that robots - just like monkeys and humans - can be intrinsically motivated to perform a series of increasingly complex tasks. By ‘learning’ from and adapting to their surroundings, the robots will develop new skills.

The project involves conducting empirical experiments with capuchin monkeys, children and adults to identify their behaviour, cognitive capacities and intrinsic motivators. Once the motivators have been identified, they will be tested on the iCub, a humanoid robotic platform and two demonstrators, Clever-K and Clever-C, which have been developed specifically for cognition experiments. The iCub is a humanoid robot which stands just over a meter high – about the size of a three-year-old child – and can manipulate objects, stand up and crawl on all fours.

According to Dr Condell, many skills not present at birth develop later in life, and the use of these robotic platforms gives a better understanding of natural and artificial cognitive systems. CLEVER-K is a technologically orientated demonstrator that will be tested in a kitchen scenario, for example developing simple tasks like sorting cultery and setting a table. CLEVER-B, will be used to reproduce and interpret the results of the experiments carried out with monkeys and children.

Dr Yianni Gatsoulis, a Research Associate in the ISRC working on the four-year project, says the project partners are confident their research will shed new light on the mechanisms underlying learning and intelligence of humans and other primates.

“The overall objective is to develop real time intrinsically motivated intelligent robots that can learn from their surroundings and adapt accordingly to carry out increasingly complex tasks. This will be an important development and one which could have a significant impact on many lives.”