Description du projet
Concevoir des robots aussi résistants que les animaux
Les robots sont l’avenir, mais malgré des décennies de recherche, ils présentent toujours un défaut majeur: ce sont des machines fragiles qui peuvent facilement tomber en panne dans des conditions difficiles. Il existe toutefois un moyen de créer des robots peu coûteux capables de se remettre de manière autonome (et immédiate) des dommages imprévus. Le projet ResiBots, financé par le Conseil européen de la recherche, révolutionnera notre approche de la tolérance aux pannes et produira des robots aussi résistants et adaptatifs que les animaux. Il utilisera notamment des algorithmes d’apprentissage par essais et erreurs qui permettent aux robots de découvrir rapidement des comportements compensatoires sans nécessiter de capteurs coûteux ou de plans d’urgence prédéfinis. L’objectif global consiste à augmenter considérablement la durée de vie des robots sans en augmenter le coût. Le projet ouvrira la voie à de nouvelles pistes de recherche pour les machines adaptatives.
Objectif
Despite over 50 years of research in robotics, most existing robots are far from being as resilient as the simplest animals: they are fragile machines that easily stop functioning in difficult conditions. The goal of this proposal is to radically change this situation by providing the algorithmic foundations for low-cost robots that can autonomously recover from unforeseen damages in a few minutes. The current approach to fault tolerance is inherited from safety-critical systems (e.g. spaceships or nuclear plants). It is inappropriate for low-cost autonomous robots because it relies on diagnostic procedures, which require expensive proprioceptive sensors, and contingency plans, which cannot cover all the possible situations that an autonomous robot can encounter. It is here contended that trial-and-error learning algorithms provide an alternate approach that does not require diagnostic, nor pre-defined contingency plans. In this project, we will develop and study a novel family of such learning algorithms that make it possible for autonomous robots to quickly discover compensatory behaviors. We will thus shed a new light on one of the most fundamental questions of robotics: how can a robot be as adaptive as an animal? The techniques developed in this project will substantially increase the lifespan of robots without increasing their cost and open new research avenues for adaptive machines.
Champ scientifique
Not validated
Not validated
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringsensorsoptical sensors
- natural sciencescomputer and information sciencesartificial intelligencemachine learningreinforcement learning
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringroboticsautonomous robots
- natural sciencesearth and related environmental sciencesphysical geographynatural disasters
Programme(s)
Thème(s)
Régime de financement
ERC-STG - Starting GrantInstitution d’accueil
78153 Le Chesnay Cedex
France