The authors set out to provide evidence that the recent introduction of transputer hardware provides sufficient computational power for online calculation of the complex algorithms needed for path planning and high-speed motion control for a robot manipulator. They present the research results in terms of performance estimates for the particular algorithms and multiprocessor architectures they chose. They present plausible arguments that it is feasible to complete the calculations identified within the specified time constraints. Unfortunately, they give almost no indication of what particular characteristics and features of the transputer are relevant.

For the reader not familiar in detail with the field, the paper also fails to make clear the significance of the results. It contains no sensitivity analysis with respect to processor speed; for example, what would the impact on the results be if the hardware were half as fast or twice as fast? Similarly, the paper gives little insight into how the particular architectures were formulated nor the criteria for the claim that they are “economically viable.” Although it is reasonably well written and interesting, the paper is more a presentation of the expert opinion of what seem to be qualified researchers than a compelling presentation of results from which the reader can reach the authors’ conclusions.