DREAM Project awarded

I was fortunate enough to be invited to the DREAM project proposal for the fiercely competitive EU H2020 call on “cognition beyond problem solving.” The proposal team was led by Stephane Doncieux from UPMC, and he did well: we were lucky enough to be awarded and are now negotiating the final stages with the EU.

So, we’ll be building dreaming robots…champagne!

The submission abstract:

A holy grail in robotics and artificial intelligence is to design a machine that can accumulate adaptations on developmental time scales of months and years. From infancy through adult- hood, such a system must continually consolidate and bootstrap its knowledge, to ensure that the learned knowledge and skills are compositional, and organized into meaningful hierarchies. Consolidation of previous experience and knowledge appears to be one of the main purposes of sleep and dreams for humans, that serve to tidy the brain by removing excess information, to recombine concepts to improve information processing, and to consolidate memory.

Our approach – Deferred Restructuring of Experience in Autonomous Machines (DREAM) – incorporates sleep and dream-like processes within a cognitive architecture. This enables an individual robot or groups of robots to consolidate their experience into more useful and generic formats, thus improving their future ability to learn and adapt. DREAM relies on Evolutionary Neurodynamic ensemble methods (Fernando et al, 2012 Frontiers in Comp Neuro; Bellas et al., IEEE-TAMD, 2010 ) as a unifying principle for discovery, optimization, re-structuring and consolidation of knowledge. This new paradigm will make the robot more autonomous in its acquisition, organization and use of knowledge and skills just as long as they comply with the satisfaction of pre-established basic motivations.

DREAM will enable robots to cope with the complexity of being an information-processing entity in domains that are open- ended both in terms of space and time. It paves the way for a new generation of robots whose existence and purpose goes far beyond the mere execution of dull tasks.

Best paper award at PPSN XIII

I’m very proud of Luis Simoes for his work on Self-Adaptive Genotype-Phenotype Maps. Our paper Self-Adaptive Genotype-Phenotype Maps: Neural Networks as a Meta-Representation was awarded best paper at PPSN XIII in Ljubljana.

In the paper, we evolve the solution as well as the genotype-phenotype mapping (through a neural network) for numerical optimisation problems. We investigate the impact of neural net settings on the locality and redundancy of the representations. Ultimately, the goal is to achieve self-adaptive mappings. The paper shows a proof-of-concept where genotype-phenotype maps successfully self-adapt, concurrently with the evolution of solutions for hard real-world problems.

Evolutionary Intelligence Journal – Special Issue on Evolutionary Robotics

Together with Nicolas Bredeche, Gusz Eiben and Stefano Nolfi, I have been fortunate to edit a special issue on Evolutionary Robotics of the Evolutionary Intelligence Journal.

If you’re interested in evolutionary robotics, check out four great papers on why the fitness function in Evolutionary Robotics has to be more than user-defined black box (Doncieux & Mouret), a method to to improve transferability between simulation and reality of co-evolving controllers and simulators (O’Dowd et al.), Evolutionary Robotics techniques to develop controllers for decentralised formation flying in space (Izzo et al.) and an investigation if t is better to learn multiple skills all at once or one at a time (Rossi & Eiben).

Evolutionary Intelligence. Volume 7 , Issue 2
Special issue on evolutionary robotics

Link: http://link.springer.com/journal/12065/7/2?wt_mc=alerts.TOCjournals

Table of contents:

Editorial (free access)
Evolutionary robotics
Evert Haasdijk , Nicolas Bredeche , Stefano Nolfi & A. E. Eiben

 

Beyond black-box optimization: a review of selective pressures for evolutionary robotics

Stephane Doncieux & Jean-Baptiste Mouret

 

The distributed co-evolution of an on-board simulator and controller for swarm robot behaviours
Paul J. O’Dowd , Matthew Studley & Alan F. T. Winfield

 

An evolutionary robotics approach for the distributed control of satellite formations

Dario Izzo , Luís F. Simões & Guido C. H. E. de Croon

 

Simultaneous versus incremental learning of multiple skills by modular robots
C. Rossi & A. E. Eiben