Simple Assembly under Visual Control
Until recently, industrial assembly has been performed either by fixed automation or by people. Fixed automation systems are in general fast, reliable, and appropriate to mass production, while manual assembly is slower but more adaptable to change. The purpose of ‘flexible automation’ is to provide an alternative to these two existing methods.
Flexible machines are intended to be modular, easily programmed to do a variety of tasks, and equipped with sensors, such as TV cameras, to observe and react to changes in their environment. The machine interprets the visual information it receives in terms of its ‘model of the world’ and instructs the mechanical actuators to make the appropriate action. In addition the results of the mechanical actions can be observed in order to improve accuracy and correct for errors.
This paper discusses some of the problems involved in constructing and using such a visually controlled machine for 3D assembly. These include manipulator control, lighting, picture acquisition, extracting three-dimensional information from two-dimensional pictures, machine calibration, system structure, and the use of multiple processors.
These problems are examined in the context of a particular class of assembly tasks, namely the vertical insertion of objects into fixtures. In particular an experimental system to perform a simple task is described in detail.
Unable to display preview. Download preview PDF.
- Saraga, P., and Skoyles, D. R., ‘An experimental visually controlled pick and place machine for industry’, 3rd International Joint Conference on Pattern Recognition, Coronado, California (November 1976).Google Scholar
- Sarage, P. and Wavish, P. R., ‘Edge tracing in binary arrays’, Machine Perception of Patterns and Pictures, Institute of Physics (Lond) Conference ser. No. 13 (1972).Google Scholar
- British Patent Application No. 7942952: An object measuring arrangement, December 1979.Google Scholar
- Habell, K. J. and Cox, A., Engineering Optics, Pitman (1953).Google Scholar
- Duda, R. O. and Hart, P. E., Pattern classification and scene analysis, Wiley Interscience (1973).Google Scholar