My work on long-term autonomy lead me to focus on the development of :
These work rely on researches related to strong AI, developmental approach and neurosciences.
The considered application frameworks - Defense, Video-games, Energy although distinct, stand on solutions based on these principles:
|The advent of “autonomous” systems (air/land/water), and their integration into the armed forces, leads to the appearance of a certain number of problems. Indeed, since communications cannot be guaranteed at all times, how can systems adapt to unforeseen situations while guaranteeing the respect of the chain of command and the rules of engagement? Our work attempts to propose an answer to these questions.||he advent of open, gigantic and persistent universes (several decades) generates a growing need for automatic generation of static content (levels, missions), but also dynamic content (non-player characters' behaviors). Indeed, from a videogame point of view, for a game to remain interesting, NPC behaviors must evolve over time according to the interactions made with the players, while remaining coherent with the elaborated universe. From a technical point of view, the consumed resources must however remain limited. Our architecture is a proposal in this direction.||Demand management and the massive arrival of renewable energies are changing the topology of the network and the roles of the various players, which are becoming dynamic (we speak of prosumer). Under these conditions, maintaining the balance of the network (injection-withdrawal) while maximizing the interests of each party requires the development of decentralized, adaptive and robust control and supervision solutions.|
|Keywords: Adaptive architecture, adjustable autonomy, decision-making under uncertainty, robustness, learning, coordination||Keywords: adaptive architecture, autonomy, decision-making, robustness, learning||Keywords: smartgrid, supervision, decentralized control, adaptive architecture, optimisation, coordination, anytime|