In the context of computer-integrated manufacture (CIM), the paper describes the need for flexible “intelligent” machinery and the need for integrated and distributed software. Methodologies in obtaining appropriate solutions are discussed and related to two major SERC sponsored research programmes at Loughborough University, which concern (i) the design of a family of mechanical and control system modules to allow robots to be configured with user defined kinematic and dynamic properties, and (ii) the design of distributed hardware and software structures, based on internationally accepted communication protocols, to allow FAS's to be produced.

In batch manufacture today's increased demand for product variety, requiring reduced lead times and work in progress, the need arises for computer-integrated manufacture (CIM). Similarly, in many conventional continuous flow manufacturing processes dramatic improvements in plant efficiency can be achieved by linking shop floor computer systems with computers performing management and design functions.

Considerable research and development work has centred on flexible manufacturing systems (FMSs)3−1 which conventionally relate to metal cutting environments. Similarly the labour intensive nature of assembly operations has led to considerable interest in the design of flexible assembly systems.3.4 The principles embodied in FMS and FAS are translatable, being applicable to most manufacturing environments.

To achieve flexible integrated manufacturing two major requirements can be identified, viz. (i) the need for flexible “intelligent” machinery, and (ii) the need for integrated but distributed software systems. This paper will consider future trends in (i) and (ii).

This report describes a new touch sensing concept in which active, exploratory touching resembles the manner in which humans use their fingertips. The concept utilizes the vibrations produced during sliding motion of the sensor to provide information for characterizing objects and features. The sensing system consists of a textured compliant artificial “skin”, a transduction element, and means for recognizing items of interest.

A relatively simple prototype was fabricated and tested for a range of potentially useful tactile tasks. Results from these experiments demonstrate the promise of the concept. The prototype sensor is shown to be effective at recognizing feature size and shape, patterns of features, feature orientation, feature position, and surface texture.

The paper deals with a three axis reaction force and displacement sensor. The design procedure of the strain gauges sensor for its intended application as a tactile proble is described. The deformable sensor body with strain gauges arrangement allows all three output voltage readings to be directly related with rectangular components of the force displacement vector in a remote contact reference system. The structural orthogonalisation that results in minimal cross component couplings is briefly outlined. The practical use of the designed probe is exemplified by seam following in a robotic welding system.

The synchronized action of robots is important in many applications, especially in the manufacturing processes. An algorithm to obtain a smooth synchronized movement of the multiple-robot system is presented for the case when only one processor is available to control all of the robots that incorporate stepper motors. The origin of this algorithm stems from the digital differential analyzer used in machine tool control.

This paper describes a new approach to obtaining a differential relationship of a robot manipulator via the Theoretical Kinematics method which may expedite computational efforts. The method involves a successive transformation of velocities from the end-effector to the base of the manipulator, link by link, using the relationship of moving coordinate systems. The equations obtained are written in the form suitable for programming on a digital computer. Furthermore, this paper also discusses the speed control model for general robot manipulators and together presents the Inverse Jacobian of cases of underdetermined and overdetermined of joint-controlled variables.

An expression is derived for the combined flexural-joint stiffness matrix and the elastic deformation field of a servo-controlled two-link robot manipulator. Such expressions are needed in dealing with light weight high-speed flexible robot manipulators. The approach employs a strain energy invariance principle with respect to the elemental and the system reference coordinate frames to derive the desired 9 × 9 combined flexural joint stiffness matrix.

For the past several years, industrial robots are being used extensively. These robots are generally equipped with relatively simple control systems. Such control systems have proved adequate, but with increased demand on robot performance, there is need for advanced and sophisticated controllers. One of the probelms in the control of robots is that system dynamics change due to several factors such as the orientation of arms and their effective inertia.

Adaptive controllers have the advantage that the system is continuously modelled and controller parameters are evaluated on-line, thus resulting in superior performance. Adaptive controllers can be realized in several ways.

This paper describes the design and performance of an explicit self tuning regulator for a robot arm.

In this paper the joint trajectories of a manipulator, which avoids obstacles in the work space and follows given path, are planned considering the dynamics of the manipulator system. The planning problem has four types of constraints: collision-free conditions, structural joints movable ranges, joints velocity limits and actuators input limits. This problem is formulated using artificial potentials which give feasible joint movements under these constraints. An algorithm using the linear programming is given to solve the problem. This algorithm enables the successive adjustment of the weighting factors of artificial potentials and gives the desired joint trajectories. The algorithm is effectively applied to the planar movements of a manipulator with four links and four degrees of freedom.