Resilient Robot Teams: a Review Integrating Decentralised Control, Change-Detection, and Learning

Robot control system using SMR signals detection

One of the important issues in designing a brain-computer interface system is to select the type of mental activity to be imagined. In some of these systems, mental activity varies with user intent and action that must be controlled by the brain-computer system, and in a number of other signals, the received signals contain the same activity-related mental activity that should be performed by t...

Behavioral control of robot teams

This paper discusses the modeling and control of fully autonomous and semi-autonomous robots, operating individually or as teams, in a behavioral context. The recent developments in networking technologies foster the use of teams of robots in a variety of applications, e.g., surveillance and rescueing. Additionally, the inclusion of human agents in a team of robots raises relevant questions on ...

Learning and Adaptation in Multi-robot Teams

An innovative decentralised motion control architecture for a humanoid robot

This paper describes the recent steps taken at PMA in the development of a pair of legs for a humanoid robot. More specifically, the paper discusses the design of a decentralised motion control architecture for the robot. Several strong arguments for hierarchical and decentralised control are given. However, robust real-time and flexible implementation of such a decentralised architecture is no...

Learning Roles: Behavioral Diversity in Robot Teams

This paper describes research investigating behavioral specialization in learning robot teams. Each agent is provided a common set of skills (motor schema-based behavioral assemblages) from which it builds a taskachieving strategy using reinforcement learning. The agents learn individually to activate particular behavioral assemblages given their current situation and a reward signal. The exper...