Although various robot behaviour learning methods have been available for some time they generally are too slow to be of much practical use. This paper provides a brief introduction to a novel robot behaviour learning method called Trajectory Velocity Learning and provides some details on implementing Trajectory Velocity Learning on sonar robots with differential drive wheels. The main advantag...

ix Chapter 1: Introduction 1 1.1 Knowledge Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2 Human-Robot Communication . . . . . . . . . . . . . . . . . . . . . . . 5 1.3 Life-Long Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.4 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.5 Outline . . . . . . . . . . . . ...

Adaptive control is challenging in real-world applications such as robotics. Learning has to be rapid enough to be performed in real time and to avoid damage to the robot. Models using linear function approximation are interesting in such tasks because they offer rapid learning and have small memory and processing requirements. This makes them suitable as adaptive controllers in nonstationary e...

This paper presents a novel machine learning method for agents of a multi-robot system. The learning process is based on knowledge discovery through continual analysis of robot sensory information. We demonstrate that classification trees and evolutionary forests may be a basis for creation of autonomous robots capable both of learning and knowledge exchange with other agents in multi-robot sys...

This paper presents a complete pipeline for learning continuous motion control policies for a mobile robot when only a non-differentiable physics simulator of robot-terrain interactions is available. The multi-modal state estimation of the robot is also complex and difficult to simulate, so we simultaneously learn a generative model which refines simulator outputs. We propose a coarse-to-fine l...

Automated methods for designing robot controllers based on machine-learning techniques have shown great promise when applied to simple robot tasks, but in order to ‘scale up’ to more complicated problems they will require assistance from human experts, a process that is often called ‘robot shaping’. In this paper, the difficult problem of learning how to visually track moving objects is examine...

This paper presents the gait pattern generation work performed for the sixlegged robot EA308 developed in our laboratory. The aim is to achieve a dynamically developing gait pattern generation structure using reinforcement learning. For the six legged robot a simplified simulative model is constructed. The algorithm constructs a radial basis function neural network (RBFNN) to command proper leg...

It currently requires years of education and practice before a skilled user can successfully program a sophisticated robot platform to perform a given task. We are exploring ways in which statistical machine learning techniques can enable Learning from Demonstration, an approach where users ‘reprogram’ a robot without writing code. In this scenario, a user demonstrates the desired task and the ...

Coordination of Communication in Robot Teams by Reinforcement Learning p. 156 Self-organized Multi-agent System for Robot Deployment in Unknown Environments p. 165 Selective Method Based on Auctions for Map Inspection by Robotic Teams p. 175 Study of a Multi-Robot Collaborative Task through Reinforcement Learning p. 185 Design of Social Agents p. 192 Event-Based System for Generation of Traffic...

This is the team description of Osaka University \Trackies" for RoboCup-99. We have worked two issues for our new team. First, we have changed our robot system from a remote controlled vehicle to a self-contained robot. The other, we have proposed a new learning method based on a Q-learning method so that a real robot can aquire a behavior by reinforcement learning.