Selected Publications:


Title:

M.S.Wilson and M.J.Neal, "Diminishing Returns of Engineering Effort in Telerobotic Systems", in IEEE Transactions on Systems, Man and Cybernetics - Part A:Systems and Humans, Special Issue on Socially Intelligent Agents - The Human in the Loop, K. Dautenhahn Ed., September 2001, volume 31, number 5, pp 459-465, IEEE Robotics and Automation Society, ISSN 0921-8890.

Abstract

Robotic systems range from teleoperated (where the human operator is in full control of all aspects of the robot's behaviour) to fully autonomous (where no human intervention takes place). The word ``telerobotic'' describes robotic systems which, although guided by a human, have a degree of autonomous behaviour.

This paper examines the trade off between the increasing design and implementation effort necessary as the system moves through the continuum from teleoperated to autonomous, and the amount of human intervention required. A case study of a human ``shepherd'' interacting with a robotic ``sheepdog'' which directs a robotic ``sheep'' is used.


Title:

M.S.Wilson and M.J.Neal, "Telerobotic Sheepdogs: how useful is autonomous behaviour?", in Proceedings of the 6th International Conference on Simulation of Adaptive Behaviour, J.A.Meyer, A.Berthoz, D.Floreano, H.L.Roitblat, and S.W.Wilson, Eds. 2000, pp 125-134, MIT Press, ISSN 1089-4365, ISBN 0-262-63200-4.

Abstract

A system modelling the interaction which takes place between a shepherd, his dog and a sheep is presented. The human shepherd interacts with a ``dog'' which is a mobile robot, which in turn controls a ``sheep'' which is a second mobile robot. This system provides a testbed for the study of how behaviour repertoire impacts on the number of interactions required to perform a task.

The robots' functionality and behaviour are presented and the commands available to the shepherd are detailed. The results of some experiments are presented which investigate the effect of introducing larger repertoires of behaviour into the dog robot. These experiments are statistically evaluated in terms of the number of interactions which are required from the shepherd with a variety of sheep displaying various patterns of behaviour.


Title:

M.S. Wilson, "Preston: A system for the evaluation of behaviour sequences," in Interdisciplinary Approaches to Robot Learning, J.Demiris and A.Birk, Eds., chapter 9. World Scientific, June 2000, Vol. 24 of Series in Robotics and Intelligent Systems. ISBN 981-02-4320-0.

Abstract

Many robot control systems have been evolved within simulated environments. Some of these have been transferred to real robots. Several have been evolved completely on real robots. Each of these methods has its drawbacks. This paper presents a system, Preston, which evolves robot controllers in the form of sequences of behaviours. These sequences are tested on a real robot and the successful results are fed back into another loop of the evolution. This paper concentrates on the development of some simple low level primitive behaviours, and how evolved sequences of these act in the real world. The repeatability of the low level behaviours is examined, along with the effect this has on the evolved behavioural sequences.

Title:

M.S. Wilson, C.M. King, and J.E. Hunt, ``Evolving hierarchical robot behaviours,'' Robotics and Autonomous Systems, vol. 22, no. 3-4, pp. 215-230, December 1997, Special Issue on Robot Learning: The New Wave, ISSN 0921-8890.

Abstract

Inspired by the work of Brooks, many researchers involved in programming robots have turned to the behaviour-based approach. At present, the behaviours are designed by hand and hard-wired into the architecture. The work presented in this paper looks at using an evolutionary algorithm approach (based on the Genetic Algorithm) to construct behaviours. Building from well defined primitive behaviours, hierarchies can be evolved to produce more complex behaviour. The behaviours in the evolutionary system are tested in simulation, but the best are then tested on a mobile robot for grounding in the real world. This allows the evolutionary process to rapidly drive the development of the behaviours using simulation while also ensuring their suitability in the real world. In the paper we show how this evolutionary process evolves practical hierarchical behaviours for the detection of a goal object in a series of mazes.

Title:

M.S. Wilson, ``Reliability and flexibility - a mutually exclusive problem for robotic assembly?,'' IEEE Transactions on Robotics and Automation, vol. 12, no. 2, pp. 343-347, April 1996, Special Issue on Assembly and Task Planning for Manufacturing, ISSN 1042-296X.

Abstract

One problem often encountered when performing assembly using robots is how to deal with variation and uncertainty in the workcell. Either the workcell becomes tightly constrained, thus losing the flexibility of the system, or errors occur, which reduces the reliability of the system. This paper attempts to address this problem using a hybrid architecture. A planning system produces a high level ordering of the assembly in terms of part motions, which are translated into robot motions by an adaptive run-time execution system. The execution system contains a flexible hierarchy of competent modular units which combine to perform the assembly reliably. The work in this paper attempts to reduce the complexity found in assembly planning and to provide a reactive run-time system which can deal with uncertainty and variation without reference to a computationally expensive global world model. The background to this work is presented, along with an experimental system designed to test out the ideas. Conclusions about the usefulness of the system are drawn at the end of the paper.