Field robotics is the use of autonomous robotic systems in highly challenging applications areas including; mining, construction, cargo handling, agriculture, subsea and aerospace systems. The focus of field robotics research is on large-scale outdoor autonomous systems in applications that are characterised by relatively unstructured, difficult and often hazardous environments. It draws together the most advanced research areas in robotics, including; navigation and control, sensing and data fusion, safety and reliability, and planning and logistics.

I. ADVANCED CYBORG (BIOLOGICAL ROBOT)

Whilst research continues to produce robots that mimic nature's creatures, scientists at the Northwestern University in Chicago, USA have developed a robot that is controlled by the brain of a sea lamprey. This is therefore an example of a biological robot, perhaps the first of many that will be developed.

Typical mobile robot structures (e.g. wheelchairs or car-like robots) do not have the required mobility for common applications such as displacement in an office, hospital, workshop, etc. New structures based on the “universal wheel” (i.e. a wheel equipped with freely rotating rollers) have been developed to increase mobility. However, these structures have important drawbacks such as spurious vibrations and limited load capacity. This paper presents a new type of omnimobile platform using an original combination of spherical wheel and universal wheel. This structure will improve the robot capabilities, i.e., load capacity and surmountable bumps.

The paper is aimed at generating optimal swing motions during the single-support phase of sagittal gait. Unlike the previous Part 1 which deals with passive motions, all joints of the biped are assumed to be active in the present Part 2. The final conditions specify an impactless heel-touch in order to avoid a destabilizing effect on the biped motion. As the biped is essentially submitted to gravity forces, the motion is generated by minimizing the joint actuating torques. Feasible motions are defined by state inequality constraints limiting joint motions, and defining foot clearance and obstacle avoidance during the swing. The optimization problem is dealt with using Pontryagin's Maximum Principle. A final two-point boundary value problem is solved by implementing a shooting method. The approach presented is illustrated by various numerical simulations applying to a seven-body planar biped which has four or five active joints during the swing phase.

E-COMMERCE AWARDS

Details of award-wining United Kingdom Information Technology projects were published by the British Computer Society's Bulletin (No. 1, 2000).

The design of prosthetic hands is constrained by a series of strict conditions. Despite this, many different design strategies have been explored. One particular form is the Southampton Hand system. This is a hierarchically controlled, electrically driven hand, with multiple axes, in an anthropomorphic form. This paper details the range of mechanical solutions adopted to address the conditions. It also compares them with other solutions.

A unified force control scheme for an autonomous underwater robotic system is proposed in this paper. This robotic system is composed of a six degree-of-freedom autonomous underwater vehicle (AUV) and a robotic arm that is mounted on the AUV. A unified force control approach, which combines impedance control with hybrid position/force control by means of fuzzy switching to perform autonomous underwater manipulation, is presented in this paper. This controller requires a dynamic model of the underwater vehicle-manipulator system. However, it does not require any model of the environment and therefore will have the potential to be useful in underwater tasks where the environment is generally unknown. The proposed approach combines the advantages of impedance control with hybrid control so that both smooth contact transition and force trajectory tracking can be achieved. In the absence of any functional autonomous underwater vehicle-manipulator system that can be used to verify the proposed controller, extensive computer simulations are performed and the results are presented in the paper.

In this paper we propose a set of techniques for a real-time motion capture of a human body. The proposed motion capture system is based on low cost accelerometers, and is capable of identifying the body configuration by extracting gravity-related terms from the sensor data. One sensor unit is composed of 3 accelerometers arranged orthogonally to each other, and is capable of identifying 2 rotating angles of joints with 2 degrees of freedom. A geometric fusion technique is applied to cope with the uncertainty of sensor data. A practical calibration technique is also proposed to handle errors in aligning the sensing axis to the coordination axis. In the case where motion acceleration is not negligible compared with gravity acceleration, a compensation technique to extract gravity acceleration from the sensor data is proposed. Experimental results not only for individual techniques but also for human motion capturing with graphics are included.

This paper describes detailed velocity kinematics and the statics of serial and parallel robots. Without resorting to the screw theory, the duality between these two types of robots is demonstrated. This duality concerns operational speeds of serial robots and operational forces of parallel robots, as well as internal speeds in redundant serial robots and internal forces in redundant parallel robots. This approach allows a deeper understanding of the duality between these two types of robots, particularly when they are redundant.

One of the simplest and natural appealing motion control strategies for robot manipulators is the PD control with feedforward compensation. Although successful experimental tests of this control scheme have been published since the beginning of the eighties, the proof of global asymptotic stability has remained unattended until now. The contribution of this paper is to prove that global asymptotic stability can be guaranteed provided that the proportional and derivative gains are adequately selected. The performance of the PD control with feedforward compensation evaluated on a two degrees-of-freedom direct-drive arm appears as fine as the classical model-based computed torque control scheme.

This paper investigates the problem of concurrent mapping and localization (CML) using forward look sonar data. Results are presented from processing of an oceanic data set from an 87 kHz US Navy forward look imaging sonar using the stochastic mapping method for CML. The goal is to detect objects on the seabed, map their locations, and concurrently compute an improved trajectory for the vehicle. The resulting trajectory is compared with position estimates computed with an inertial navigation system and Doppler velocity sonar. The results demonstrate the potential of concurrent mapping and localization algorithms to satisfy the navigation requirements of undersea vehicles equipped with forward look sonar.

This paper describes a path planning method for a small autonomous aerial vehicle to be used for inspecting overhead electricity power lines. A computational algorithm is described which converts a standard three dimensional array representation of one or more obstacles in the vehicle's environment into an octree and a connectivity graph. This achieves a significant reduction in computer memory usage and an increase in execution speed when the graph is searched. Path planning is based on a three-dimensional extension of the distance transform. Test results demonstrate rapid and effective operation of the planner within different workspaces.

This paper analyzes the dynamics and control of pinch motions generated by a pair of two multi-degrees-of-freedom robot fingers with soft and deformable tips pinching a rigid object. It is shown firstly that passivity analysis leads to an effective design of a feedback control signal that realizes dynamic stable pinching (grasping), even if extra terms of Lagrange's multipliers arise from holonomic constraints of tight area-contacts between soft finger-tips and surfaces of the rigid object and exert torques and forces on the dynamics. It is shown secondly that a principle of superposition is applicable to the design of additional feedback signals for controlling both the posture (rotational angle) and position (some of task coordinates of the mass center) of the object provided that the number of degrees of freedom of each finger is specified for satisfying a condition of stationary resolution of controlled position state variables. The details of feedback signals are presented in the case of a special setup consisting of two robot fingers with two degrees of freedom.

This paper describes the autonomous navigation and control of an undersea vehicle using a vehicle control architecture based on the Distributed Architeclure for Mobile Navigation and a terrain-aided navigation technique based on simultaneous localisation and map building. Development of the low-speed platform models for vehicle control and the theoretical and practical details of mapping and position estimation using sonar are provided. Details of an implementation of these techniques on a small submersible vehicle “Oberon” are presented.

In this paper fhe problem of motion control of a biped is considered. We develop a new method based on multi-agent associated Neural AIGLS (On-line Augmented Integration of Gradient and Last Sguare method) – RSPN (Recursive Stochastic Petri Nets) strategy. This method deals with organization and coordination aspects in an intelligent modeling of human motion. We propose a cooperative multi-agent model. Based on this model, we develop a control kernel named IMCOK (Intelligent Motion COntrol Kernel) which consists of a controller, a coordinator and an executor of different cycles of the motion of the biped. When walking, IMCOK receives messages and sends offers. A Decision Making of Actions (DMA) is developed at the supervisor level. The articulator agents partially planify the motion of the associated non-articulator agents. The system is hybrid and distributed functionally. The learning of the biped is performed using an On-line Augmented Integration of Gradient and Last Sguare Neural Networks based algorithm. In the conflictual situations of sending or receiving messages by the managers of MABS we apply a new strategy: Recursive Stochastic Petri Nets (RSPN). This module is fundamental in the On-line information processing between agents. It allows particularly the Recursive strategy concept. Cognitive agents communicate with reactive (non-articulator) agents in order to generate the motion.

A work cell with multiple robots increases manufacturing flexibility and productivity. Robots in the work cell equipped in a sparse area usually share a motion path, resources and workspace. In this paper, an embedded Markov chain model for a multi-robot system that has a common workspace is constructed on the basis of the concept of a multi-processor system. With the presented model, we measure the performance of error recovery schemes under different workloads and analyze the sensitivity of the execution time with respect to the robot speed. We verify the presented model with an experimental multi-robot work cell. This study is useful in evaluating the performance of a robotic work cell and presents a guide for designing a complex work cell.

This paper describes a newly developed algorithm for CAD-based dimensional inspection path planning utilizing coordinate measuring machines (CMMs). The algorithm guarantees to provide optimum collision-free inspection probe paths by using the topological structure of the boundary representation (B-rep) solid CAD models of the inspected parts. A concept of automatic generation of optimum and collision free path in three dimensional space using available CAD database is proposed. The algorithm is generic for generating probe path over prismatic polyhedral parts. It serves as a principal part of the inspection path planning system. It is based on the modified ray tracing technique which uses a B-Rep data from any geometric modeling systems. Between start point and target point, an imaginary ray is established and if an intersection with part is encountered, an optimal detour path is created avoiding interference of the probe with the part. The generated path consists of linear segments joining start point and target point by various intermediate points. To locate these intermediate points in the consideration space, topological and geometrical structures of the part models are used at the time of decision making. To examine and implement the algorithm, a user-friendly application is developed employing AutoCAD Runtime Extension (ARX) development environment with object oriented programming (OOP) techniques, running on a Windows NT workstation. The effectiveness of the proposed algorithm is verified by the results of the implementation demonstrating optimum collision-free dimensional inspection path generation for four representative prismatic part models. All in all, this work contributes to the knowledge-base formation of automated dimensional inspection research area and paves a way for the integration of CMMs into a CAD/CAM environment, thus automate the process of design, manufacturing and quality assurance.

It is known that it is impossible to select fixed gains for a PD controller that will critically damp the response to disturbances for all configurations of a given robot system. Because of this the potential for overshoot is always present and cannot be avoided unless the system is severely overdamped. This is not necessarily a practical solution and can be an economically unacceptable approach. On the other hand, however, if overshoot is permissible to some degree for some systems in the case of conventional Serial robots it is still prohibited in the case of Parallel robots as it may easily bring the robot to one of its possible singular configurations, causing damage to the system. This paper introduces a new algorithm for the design of PD controllers that ensures uniform and fast dynamic responses, which are free from overshoots for all robot configurations. The technique also satisfies general stability requirements for the system.

Near minimum-time direct voltage control (DVC) algorithms synthesizing path-planning and path-following are proposed for wheeled mobile robots (WMRs) satisfying (i) initial and final postures and velocities as well as (ii) voltage and current constraints. To overcome nonholonomic and nonlinear properties of WMRs, we divide our control algorithm for cornering motion into three sections: TSD (Translational Section of Deceleration), RS (Rotational Section), and TSA (Translational Section of Acceleration). We developed off-line DVC algorithms using the quadratic problem with the object function minimizing the total time, where voltages to the motors are controlled directly without velocity/torque-servo modules, while satisfying the current constraints. Two methods of searching for the two control parameters (number of steps for RS MR and velocity constraint in RS β) were considered: The one is composed of one simple 1-dimensional search for β, and the other is composed of two 1-dimensional searches for MR and β which has better performance. Performances of the proposed control algorithms are validated via various simulations.

A fast and accurate method for detecting the collisions of convex polyhedra in a graphical simulation environment based on a newly developed method of distance estimate is presented. By the simultaneous use of the enclosing and the enclosed ellipsoids of convex polyhedra, potential collisions can be detected more accurate than those methods using only bounding volume for object representation, and more efficient than the polyhedral methods. An approach for computing the enclosed ellipsoid of a convex polyhedron by compressing, stretching and scaling operations on its best-fit enclosing ellipsoid is introduced. Graphical simulations of two case studies (moving polyhedral objects in three-dimensional space and multiple robot arms undergoing straight line motions) are conducted to demonstrate the accuracy of the proposed algorithm for collision detection.