Paul Chernett

Automating agricultural vehicles

Describes the design and construction of TROWEL ‐ a test bed for experimental agricultural vehicles. The vehicle will be used to explore ways of increasing the productivity of expensive agricultural mobile machinery by taking over some of the tasks of the operator, allowing him to drive faster or for longer; and by allowing a single operator to control several machines simultaneously. In some cases machines may be able to operate entirely autonomously without operator intervention.

Mixing simulations and real subsystems for subsea robot development. Specification and development of the core simulation engine

The core simulation engine is a distributed hardware and software environment that allows the interoperable use of simulations and real subsystems. Its primary role is to allow incremental, remote integration and testing of complex engineering systems such as underwater vehicles. Further applications include operator training, mission feasibility assessment and post mission replay. The paper discusses key issues in developing such a hardware-in-the-loop-simulation (HILS) involving an environment and physically distributed sub-systems operating in different logical and real-time modes. Time synchronisation, interoperability, simulation/real world calibration, and a subscription communications protocol are discussed. An implementation using PVM for distributed simulation support, dVise for modelling and a DSP-based vehicle communication spine is presented. A spin off approach for improved UUV navigation is also described.

HIERARCHICAL BEHAVIOURAL CONTROL FOR AUTONOMOUS VEHICLES

Abstract The subject of this paper is the control of autonomous vehicles. A hierarchical approach is studied in the context of fuzzy systems and a programming language for the mid to low level control of autonomous vehicles is described. The language, called FDTL (Fuzzy Decision Tree Language), is based on a computational model that combines fuzzy rule based control with the hierarchical nature of decision trees.

A Robust Navigation Model for an Autonomous Mobile Robot

This paper reports on preliminary work with an adaptive multi-representational approach to modelling the real world. The proposed method operates in real time, accounts for low accuracy in the robots sensors, requires minimal computational resources and leads to the implementation of efficient path planning algorithms. Experiments are currently being conducted on a mobile indoor robot. In the longer term, this work will be extended to an outdoor farm vehicle.

SPREAD: A Distributed Simulation TOOLKIT

This article describes “SPREAD”, a simulation tool kit and its use in building “Virtual Robots”, a simulation of multiple mobile robot vehicles used in the teaching of Computer Science at university level. A novel aspect of the simulator is the use of PVM [1] to achieve high performance at low cost by using spare CPU cycles on large numbers of networked workstations. Acknowledgments Many people have been involved in developing the simulator. In particular our thanks go to Paul Marriot who wrote the very first version as long ago as 1992, to Caleb Ying who was responsible for many of the architectural decisions, and to Christos Voudouris for many ideas and clarifications that have found their way into the programs.

A Network-Centric Approach to Computer Science Education and Research Based on Robotics

Summary A network-centric university teaching and research environment based around mobile robots is described. It supports teaching a wide spectrum of computer science and engineering subjects (eg digital electronics, computer architecture, operating systems, communications, distributed processing, software engineering, machine vision, artificial intelligence, embedded computer design, real-time computing, HCI, Computer Graphics, Virtual Reality etc) within a single unified experimental environment. This enables students to combine their coursework so as to build more complex, interesting and industrially relevant systems thereby significantly increasing their motivation and knowledge. The laboratory infrastructure makes extensive use of networks to produce a versatile and cost effectiveness teaching and research environment. This paper will focus on the teaching aspects of the laboratory describing the technical infrastructure showing how it is used to support the Essex University information technology curriculum and report on the effectiveness of the approach. Finally, to demonstrate the more general academic synergy provided by this laboratory a short overview of some of the research applications is presented.

PROCESSING MODULES FOR HYBRID, REAL-TIME,DISTRIBUTED ROBOT ARCHITECTURES

New experimental hardware for research into architectures for distributed intelligent embedded systems is proposed that will provide a wide range of communication media including non- deterministic broadcast such as Ethernet, deterministic broadcast such as CAN and processor busses such as VME. The emphasis is on large scale system integration rather than provision for individual capabilities. A prototype implementation of some of the proposed hardware and software modules is described together with their use in several research projects.