Dr Michael Dickinson

The paper describes a technique to determine and to track the three-dimensional 'world' coordinates of a point in space, relative to some defined datum. The work was initiated in order to track the location of a robot manipulator end effector, but has relevance to automatic guided vehicles and other applications. A prototype system has been built and evaluated, and has been used successfully to track the tip of a robot manipulator arm moving at velocities of up to 2 m/s within a working envelope consisting of a cube of side 2 m. Measurement resolution of 0. 5 mm was obtained, and a sampling rate of 300 Hz was achieved, by the use of real-time software within the measurement system. The coordinates and the velocity outputs are computed in only 1. 2 ms. The technique provides accurate tracking of end-effector position, which can be used both for evaluating the dynamic performance of existing controllers and for the synthesis of new forms of control.

At present, knowledge of robot end-effector position in world (spatial) coordinates is derived from measurements of robot joint positions, using the mathematical relationship between joint and end-effector positions. The alternative solution proposed in this paper is the use of an ultrasonic position measurement and tracking system. In this system, ultrasonic receivers are mounted on the vehicle and rotated by a motor drive. Measurement of the phase differences between the signals received yields the range and bearing of the transmitter. However, measurement errors of 50mm were reported and computation times of 90ms per measurement were required, and hence such a system is unable to satisfy the current requirement.

Statistical process control is the application of statistics to the monitoring and control of industrial processes. Measures of process variables are taken usually at regular intervals; these measures may be estimates of the value of the variable and the range of spread of the measure. Process trends can be observed; long term variation in the monitoring variable would indicate, for example, changes in process parameters while short term discontinuities could indicate changes in material quality. Statistical Process Control (SPC) techniques have recently become more important to industry, particularly with the increasing need to monitor processes in order to improve quality of the product, to reduce operating costs and to optimise the process. This paper considers the application of SPC techniques using a personal computer for data capture, processing and storage. It is the result of a study into the possible application of SPC to the manufacture of lamp caps.

This paper shows how personal computer systems have been applied in one industry, and in each of three applications has provided essential quality control information, as well as development capability for future improvement. The three applications considered are from the automobile industry, and in each the computer is used purely to monitor a process. The three applications are: overdrive clutch test rig, impact testing system for car bumpers, and road simulator. The paper shows how hardware is selected for the application, and the software may then be developed to produce appropriate output of data in an understandable form on suitable peripherals.