The University of Sheffield
Department of Psychology

21 July 2008

Departmental Researchers receive funding boost for robot learning project

Kevin Gurney, Pete Redgrave, and Tom Stafford have been awarded FP7 (Seventh Research Framework Programme) funding for a project entitled "Intrinsically Motivated Cumulative Learning Versatile Robots" (IM-CLeVeR). The research is led by Gianluca Baldassarre at ISTC (Institute of Cognitive Sciences and Technologies), Italy.

Negotiations still have to take place with the European Union on budgetary details but the expected award is just over €700,000. This will fund two postdoctorals in the Department working on intrinsically motivated learning: a computational modeller, and another doing experimental work with humans. The project is due to start in January 2009. An abstract for the project is below.

Abstract:
IM-CLeVeR aims to develop a new methodology for designing robots controllers that can:


  2. cumulatively learn new efficient skills through autonomous development based on intrinsic motivations, and
  3. reuse such skills for accomplishing multiple, complex, and externally-assigned tasks.

During skill-acquisition, the robots will behave like children at play which acquire skills autonomously on the basis of “intrinsic motivations”. 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 readapting previously acquired skills. This overall goal will be pursued investigating three fundamental scientific and technological issues:


  2. the mechanisms of abstraction of sensory information;
  3. the mechanisms underlying intrinsic motivations, e.g. “curiosity drives” that learn to focus attention and learning capabilities on “zones of proximal development”;
  4. hierarchical recursive architectures which permit cumulative learning.

The study of these issues will also be fuelled by a reverse-engineering effort aiming at reproducing with bio-mimetic models the results of empirical experiments run with monkeys, children, and human adults. The controllers proposed will be validated with challenging demonstrators based on a single humanoid robotic platform. As a main outcome, the project will produce core scientific and technological knowledge on the basis of the integrated work of a highly interdisciplinary Consortium involving leading international neuroscientists, psychologists, roboticists and machine-learning researchers. Overall, the project will significantly advance the scientific and technological state of the art in autonomous learning systems, both in terms of theory and implementations. It will therefore provide key foundation stones for the development of the next generation of autonomous learning robots.