ECAI 2004 Conference Paper

[full paper]

Robel : Synthesizing and Controlling Robust Robot Behaviors

Benoit Morisset, Guillaume Infantes, Malik Ghallab, Felix Ingrand

We present the Robel supervision system which is able to learn from experience robust ways of performing high level tasks (such as "navigate to"). Each possible way to perform the task is modeled as a Hierarchical Tasks Network (HTN), called modality whose primitives are sensory-motor functions. An HTN planning process synthesizes all the consistent modalities to achieve a task. The relationship between supervision states and the appropriate modality is learned through experience as a Markov Decision Process (MDP) which provides a general policy for the task. This MDP is independent of the environment and characterizes the robot abilities for the task.

Keywords: robotics, supervision, learning, robustness, planning, navigation, Markov Decision Process, Hierarchical Task Network

Citation: Benoit Morisset, Guillaume Infantes, Malik Ghallab, Felix Ingrand: Robel : Synthesizing and Controlling Robust Robot Behaviors. In R.López de Mántaras and L.Saitta (eds.): ECAI2004, Proceedings of the 16th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2004, pp.1067-1068.