Some speculations on feature recognition

It is a paradox that today's automated systems for geometry-based engineering activities such as process planning are not able to use geometric information directly. While geometric models have been available for some time, it is not proving easy to link them to decision-making systems. The main hope for building this bridge is feature recognition, in which the geometry of an object is interpreted in terms of geometric elements of engineering significance. In this speculative paper, some of the problems of applying such a process to shape models are outlined, and some lines of advance are suggested. In particular, the possible use of the set-theoretic (constructive solid geometry (CSG)) solid model for feature recognition is discussed. © 1988.