Qualitative physics is a very interesting combination of Artificial Intelligence methods and Physics and conceptually in many parts not far away from applications and formalizations using fuzzy logics. This book especially deals with the problems of spatial reasoning and planning which are especially important for design stages of mechanical systems where system parameters have not yet been specified or the problems have been specified only on a qualitative level. The main application area is the computer-based design of mechanisms and robots, an area that in classical physics is highly dominated by differential equations and numerical methods. The approach presented in this book complements the existing methods and allows the immediate derivation of qualitative results and the planning of paths. The central issues of the book are as follows: ‒ How to map continuous spatial quantities into qualitative representations that enable inferences about qualitative spatial relationships and the synthesis of a sequence of qualitative spatial configurations. ‒ How to approximately describe the constrained-motion path and the constrained velocity relationships of a mechanism. ‒ How to derive the path of a mechanism which geometrically satisfies the constraints imposed by its joints and by its environment. The presentation of the material is very interesting and will be an excellent example for further work at similar problems. It starts with the representation of qualitative spatial relationships, especially a qualitative distance and qualitative angles, concepts that can be understood as elements of a fuzzy geometry. Based on these concepts, a methodology of spatial reasoning and planning can be developed that uses, for instance, qualitative trigonometry and qualitative arithmetic. These formalisms then can be applied to the qualitative and quantitative analysis of configurations. The reasoning about velocity relationships is dealt with in the next steps. The combination of the developed theoretical background allows the solution of several applied questions, such as the qualitative planning of robot motions, graphical simulations and the generation of spatial maps, which is illustrated by some experiments and examples. It is very interesting to see the approach and the results achieved so far. It can be seen that in many disciplines the classical approaches (mainly based on the combination of mathematics and the problem area, mechanics in this case) get a third component based on computer-oriented AI-methods, and a lot of progress can be expected from this threefold combination. It also indicates the steps to be introduced into the education of engineers and scientists in many areas. This book can be used for graduate students in several areas, but also (possibly even more important) for the continuous education of people that are already working in a given field.
Reviewer: Christian Posthoff (St. Augustine)