Derek W. Seward

The Development, Control and Operation of an Autonomous Robotic Excavator

The excavation of foundations, general earthworks and earth removal tasks are activities which involve the machine operator in a series of repetitive operations, suggesting opportunities for the automation through the introduction of robotic technologies with subsequent improvements in machine utilisation and throughput. The automation of the earth removal process is also likely to provide a number of other benefits such as a reduced dependence on operator skills and a lower operator work load, both of which might be expected to contribute to improvements in quality and, in particular, the removal of the need for a local operator when working in hazardous environments.The Lancaster University Computerised Intelligent Excavator or LUCIE has demonstrated the achievement of automated and robotic excavation through the implementation of an integrated, real-time, artificial intelligence based control system utilising a novel form of motion control strategy for movement of the excavator bucket through ground. Having its origins in the systematic observation of a range of machine operators of differing levels of expertise, the control strategy as evolved enables the autonomous excavation of a high quality rectangular trench in a wide variety of types and conditions of ground and the autonomous removal of obstacles such as boulders along the line of that trench.The paper considers the development of the LUCIE programme since its inception and sets out in terms of the machine kinematics the evolution and development of the real-time control strategy from an implementation on a one-fifth scale model of a back-hoe arm to a full working system on a JCB801 360° tracked excavator.

Development of a Multi-Arm Mobile Robot for Nuclear Decommissioning Tasks

This paper concerns the design of a two-arm mobile delivery platform for application within nuclear decommissioning tasks. The adoption of the human arm as a model of manoeuvrability, scale and dexterity is the starting point for operation of two seven-function arms within the context of nuclear decommissioning tasks, the selection of hardware and its integration, and the development of suitable control methods. The forward and inverse kinematics for the manipulators are derived and the proposed software architecture identified to control the movements of the arm joints and the performance of selected decommissioning tasks. We discuss the adoption of a BROKK demolition machine as a mobile platform and the integration with its hydraulic system to operate the two seven-function manipulators separately. The paper examines the modelling and development of a real-time control method using Proportional-Integral-Derivative (PID) and Proportional-Integral-Plus (PIP) control algorithms in the host computer with National Instruments functions and tools to control the manipulators and obtain feedback through wireless communication. Finally we consider the application of a third party device, such as a personal mobile phone, and its interface with LabVIEW software in order to operate the robot arms remotely.

Artificial intelligence in the control and operation of construction plant - the autonomous robot excavator.

Abstract The paper describes a significant development in the intelligent automatic control of an adaptive robot. The particular application concerns the problem of autonomous excavation, and the subtlety of the approach lies in the fact that the machine can cope with highly variable ground conditions and even underground obstructions without human interference. The nature of the excavation problem is first described. The development of a working fifth-scale model of an excavator—LUCIE—the Lancaster University Computerised Intelligent Excavator. The hardware systems are briefly described, but the main emphasis of the paper concerns the method by which the AI technique of a rule-based “production system” is used to control the tip of the excavator bucket. The software is in two distinct parts; a low-level velocity controller and a high level “activities manager” which contains the rules. It is concluded that the best way to achieve subtle, human-like, automatic control is by means of simple rules and heuristics, and not complex control algorithms. Finally the paper reports field trials of LUCIE, and goes on to describe proposed future developments.

The Development of Research Models for Automatic Excavation

Lancaster University has recently begun research into construction robotics and initial concentration is on excavation plant. This paper describes the building of computer and small-scale models to facilitate research and development into hardware and control strategies. The emphasis is on maintaining flexibility to allow the research the widest possible scope. For this reason a modular system for the models was devised so that compatible components can built-up in various ways. A selection of these modules is described. 1. Background A multi-disciplinary team of civil, electrical and mechanical engineers has recently established construction robotics research at Lancaster. Initially it is the intention of the group to concentrate on the excavation process, and the back-hoe excavator arm provides an excellent vehicle for exploring the required technologies. Existing back-hoe arms have already evolved into efficient structures of robotic form, and there is ample scope for adding intelligence in order to improve speed of operation, accuracy, fuel economy, ease of use and independence from conventional setting-out procedures. Existing excavator arms weigh about one tonne travel at 3 metres per second and develop ram forces of up to 20 tonnes. Although it is intended to eventually concentrate the work on a real back-hoe, it was decided that for reasons of safety and convenience it is highly desirable to be able to experiment with control strategies firstly on a computer simulation and secondly on a small-scale model where the forces involved have less potential for damage. In designing the models the emphasis was placed on flexibility and compatibility, and for this reason a modular approach was adopted. A selection of possible modules is shown in Figure 1 and although it may be decided not to proceed with all the modules, the system is designed to keep all options open at the outset. 2. Computer Simulation In working with the model arm, computer simulations will be used to investigate, develop and establish operational and control strategies which can then be incorporated into the model. Initially, this will require that the general operational environment into which the robot arm is to be introduced should be modelled and the operation of the arm within that. environment considered. Once a suitable model of the environment

LUCIE the robot excavator-design for system safety

Staff and students at Lancaster University have, for the past five years, been involved in the development of an autonomous robot excavator LUCIE (Lancaster University Computerised Intelligent Excavator). The aim of the project is to add autonomy in order to produce a robot excavator with the following characteristics: it should concentrate on the task of trenching; it should adapt to different soil types without human intervention; it should cope with obstructions; and it should eventually be a self-contained system with no cables to external computers. The hardware and software architectures, optical distance sensor, and safety problems are described.

Controlling an intelligent excavator for autonomous digging in difficult ground.

This paper reports on the recent advances made in developing an autonomous robot excavator. Previous work on a fifth-scale model was reported at earlier symposia, but the technology has now been transferred to a real excavator LUCIE Lancaster University Computerised Intelligent Excavator. The paper concentrates on the architecture of the software control which enables the machine to modify its behaviour to cope with highly varying ground conditions. A single powerful processor controls the low-level motion of the excavator arm, as well as high level tactical and strategic behaviour. An A.I. rule-based approach has been adopted for high-level functions. Successful field trials are reported.

Safety analysis of autonomous excavator functionality.

This paper presents an account of carrying out a hazard analysis to define the safety requirements for an autonomous robotic excavator. The work is also relevant to the growing generic class of heavy automated mobile machinery. An overview of the excavator design is provided and the concept of a safety manager is introduced. The safety manager is an autonomous module responsible for all aspects of system operational safety, and is central to the control systems architecture. Each stage of the hazard analysis is described, i.e. system model creation, hazard definition and hazard analysis. Analysis at an early stage of the design process, and on a system that interfaces directly to an unstructured environment, exposes certain issues relevant to the application of current hazard analysis methods. The approach taken in the analysis is described. Finally, it is explained how the results of the hazard analysis have influenced system design, in particular, safety manager specifications. Conclusions are then drawn about the applicability of hazard analysis of requirements in general, and suggestions are made as to how the approach can be taken further

The anatomy of a humanoid robot

This paper investigates the feasibility of constructing a humanoid robot using existing technology. Firstly, the adoption of the humanoid form is justified. The structure, strength and power capabilities of a human are analysed in engineering terms, and taken to represent the requirements specification for a humanoid robot. Technological alternatives to the biological components are reviewed and compared to this specification. The feasibility of matching human performance is considered, and it is concluded that the necessary power and energy requirements can be fitted within the mass and volume of the human body.

Developing real-time autonomous excavation-the LUCIE story

Earthworks and earth removal are activities which involve a series of repetitive operations on the part of the machine operator, suggesting opportunities for the automation of these processes with resultant improvements in machine utilisation and throughput. The automation of the earth removal process also provides a number of secondary benefits such as a reduced dependence on operator skills and a lower operator work load. The achievement of automated and robotic excavation in the form of the Lancaster University Computerised Intelligent Excavator or LUCIE was dependent on the implementation of a real-time artificial intelligence based control system utilising a novel form of motion control strategy for movement of the excavator through ground. Based on the systematic observation of a range of machine operators, the control strategy evolved permits the autonomous excavation of a high quality rectangular trench in a variety of types of ground and of the removal of obstacles along the line of that trench.

The automation of piling rig positioning using satellite GPS

The paper is divided in two parts. Part one describes the Stent Automatic Pile Positioning and Recording system (SAPPAR) which was launched in November 1994. The system utilises a Trimble satellite global positioning system (GPS) to assist rig drivers in accurately positioning the rig over a pile position without the need for setting out. Advantages of the system include: cost savings by removing the need for site survey staff; faster set-up times over pile positions; increased accuracy — the system can reliably position the rig to within ± 25 mm; removal of problems resulting from damage to setting out pins; constant monitoring of pile position; and Links to CAD for data input and as-built drawings. Part two describes a further development of the system in collaboration with Lancaster University and Casagrande, the Italian rig manufacturer. The aim of the research is to fully automate the final positioning process. This represents one of the first uses of GPS for real-time automation. The system hardware components include: ultra-compact PC1O4 processor cards for a compact and robust embedded system; minimum sensing on the rig to minimise cost and maximise robustness; and limit sensors to facilitate on-board safety. The control algorithms were developed on a fifth-scale model in the laboratory using an innovative and new approach to the design of model based control systems. The importance of careful consideration of safety issues is stressed and conclusions are drawn based on the early findings from preliminary field trials.