This thesis addresses situated, embodied agents interacting in complex domains. It focuses on two problems: 1) synthesis and analysis of intelligent group behavior, and 2) learning in complex group environments. Basic behaviors, control laws that cluster constraints to achieve particular goals and have the appropriate compositional properties, are proposed as effective primitives for control and learning. The thesis describes the process of selecting such basic behaviors, formally specifying them, algorithmically implementing them, and empirically evaluating them. All of the proposed ideas are validated with a group of up to 20 mobile robots using a basic behavior set consisting of: safe--wandering, following, aggregation, dispersion, and homing. The set of basic behaviors acts as a substrate for achieving more complex high--level goals and tasks. Two behavior combination operators are introduced, and verified by combining subsets of the above basic behavior set to implement collective flocking, foraging, and docking. A methodology is introduced for automatically constructing higher--level behaviors

This paper describes the problem of task sharing between two autonomous six--legged robots. The robots are equipped with two-way communication, object and goal sensing, and a repertoire of basic behaviors. The performance on the selected task, pushing an elongated box toward a goal region, is difficult for a single robot and improves significantly when performed cooperatively. The cooperative solution, however, requires careful coordination between the robots. We present an approach that takes advantage of a simple communication protocol to compensate for the robots' noisy and uncertain sensing and actuation, and their partial knowledge about the world. We demonstrate our approach on a series of experiments with two physical robots. 1 Introduction The work in this paper is concerned with the problem of mobile robot task sharing. In particular, we are focusing on manipulation tasks with sufficiently challenging dynamics to require the careful cooperation of two (or more) robots. In m...

In this paper, we examine the problem of controlling multiple behaviour-based autonomous robots. Based on observations made from the study of social insects, we propose ve simple mechanisms used to invoke group behaviour in simple sensor-based mobile robots. The proposed mechanisms allow populations of behaviour-based robots to perform tasks without centralized control or use of explicit communication. We have veri ed our collective control strategies by designing a robot population simulator called SimbotCity. Wehave also constructed a system of ve homogeneous sensor-based mobile robots, capable of achieving simple collective tasks, to demonstrate the feasibility of some of the control mechanisms.

Abstract not found

Abstract not found

In this paper, we explore the idea of using an Adaptive Logic Network (ALN) [2] for behaviour arbitration. Our approach is to define the collective task, to be performed by multiple robots, as a group behaviour. The group behaviour is a set of behaviours, each of which specifies a single step in the collective task. An environmental cue can be used to control the transition between behaviours, thus allowing the progress of the collective task to self-govern its execution. We simplify the behaviour arbitration in the collective task by training an ALN implemented using simple combinational logic. We provide a description of our Collective Robotic Intelligence Project (CRIP) including our simulation results and our multi-robot system on which these results will be deployed. 1 Introduction Can collective tasks be accomplished using group behaviours ? Interest in accomplishing tasks by using multiple robots has resulted in systems designed using cooperative behaviour [11, 8, 1, 7, 17]. Ou...

Does coherent behaviour require an explicit mechanism of cooperation? In this dissertation, the relationship between local perception and global action in a system of multiple mobile robots was examined for a collective box-pushing task. The problem investigated was how local sensing could be used to coordinate the individual motor responses of a system of robots in a coherent manner, using only implicit communication through the task. The task was to move a large box from an initially unknown position to a specified goal location. The central thesis put forward, is that for the box-pushing task a coherent behaviour is possible, without an explicit mechanism of cooperation, by using the mass effect of a system of redundant robots. Preliminary work in collective robotics appeared to lend weight to the hypothesis that collective tasks, by multi-robot systems, are possible without centralized control or explicit inter-robot communications, two common control mechanisms used for cooperati...

We are experimenting with multiple mobile robots under visual control. The current goal of our work is to construct an architecture for robots engaged in cooperative and competitive behaviour. Multiple robots have recently been the focus of much attention; however, the little work that has been done with implemented systems has involved simple tasks and behaviours. We have initiated the Dynamo (Dynamics and Mobile Robots) Project to provide a link between theory and practice. One part of this project is the Dynamite testbed which consists of a fleet of radio controlled vehicles that receive commands from a remote computer. Robot position and orientation is determined using off-board visual sensing. We have chosen soccer playing as a domain for our experiments since it requires real-time interaction with a dynamic environment. It involves inter-robot cooperation as well as competition between teams. We outline two complementary approaches taken in our laboratory to robot control. The first, Constraint Nets, is a model for robotic systems and behaviours, which provides a theoretical foundation for systems design and analysis.