Abstract

This paper describes a formulation of reinforcement learning that enables learning in noisy, dynamic environemnts such as in the complex concurrent multi-robot learning domain. The methodology involves minimizing the learning space through the use behaviors and conditions, and dealing with the credit assignment problem through shaped reinforcement in the form of heterogeneous reinforcement functions and progress estimators. We experimentally validate the approach on a group of four mobile robots learning a foraging task. 1 Introduction Developing effective methods for real-time learning has been an on-going challenge in autonomous agent research and is being explored in the mobile robot domain. In the last decade, reinforcement learning (RL), a class of approaches in which the agent learns based on reward and punishment it receives from the environment, has become the methodology of choice for learning in a variety of domains, including robotics. In this paper we describe a formulat...

Citations

2441	A robust layered control system for a mobile robot - Brooks - 1986
1060	Learning to predict by the methods of temporal differences - Sutton - 1988
472	Learning to act using real-time dynamic programming - Barto, Bradtke, et al. - 1995
316	Automatic programming of behavior-based robots using reinforcement learning - Mahadevan, Connell - 1992
280	Learning in embedded systems - Kaelbling - 1993
228	Neural Network Perception for Mobile Robot Guidance - Pomerleau - 1992
220	Multi-agent reinforcement learning: independent vs. cooperative agents - Tan - 1993
190	Learning to coordinate behaviors - Maes, Brooks - 1990
147	Learning to predict by the methods of temporal di erence - Sutton - 1988
139	Interaction and intelligent behavior - MATARIC - 1995
125	Designing emergent behaviors: from local interactions to collective intelligence - Mataric - 1993
102	Heterogeneous Multi-Robot Cooperation - Parker - 1994
93	The behavior language user’s guide - Brooks - 1990
64	Learning to behave socially - Mataric - 1994
61	Programming robots using reinforcement learning and teaching - Lin - 1991
48	Using local models to control movement - Atkeson - 1989
39	Scaling reinforcement learning to robotics by exploiting the subsumption architecture - Mahadevan, Connell - 1991
35	Integrating inductive neural network learning and explanation-based learning - Thrun, Mitchell - 1993
32	Self-improving reactive agents:: Case studies of reinforcement learning frameworks - Lin - 1990
27	Fast, robust adaptive control by learning only forward models - Moore - 1992
16	Behavior-Based Systems: Key Properties and Implications - Mataric - 1992
14	Kin Recognition, Similarity, and Group Behavior - Mataric - 1993
6	Model-Based Robot Learning - Atkeson, Aboaf, et al. - 1988
6	Learning reactive sequences from basic reflexes - Millan - 1994
3	Coordination of Multiple Behaviors Acquired by A Avision-Based Reinforcement - Asada, Uchibe, et al. - 1994
2	Robot Juggling: An Implementation of Memory-Bassed Learning - Atkeson - 1994