Qualitative physics is a field of research that provides a new repertoire of reasoning techniques for computers. These techniques enable computers to reason about the real world without using numerical information or methods. The aims of qualitative physics are related to those of ``Naive Physics'', which attempts to describe the world using a formalised version of human non-specialist understanding, thereby developing a model of human reasoning and also providing commonsense knowledge for A.I. systems. Qualitative physics tends to be applied to more restricted problem domains than naive physics - particularly engineering problem domains such as hydraulics, formal mechanics and electronics, rather than to problems in the everyday world.
Qualitative physics research to date has concentrated on problems that do not involve much spatial information, dealing instead with complex processes or device interaction over time. The analysis of systems such as circuits can be performed by using a simple representation of the connections between components in the circuit, rather than any spatial description of the circuit components.
Early qualitative reasoning systems operated in spatial problem domains, but used methods of reducing the spatial content in the problem until the programs operated only on connectivity information. The ways in which this was achieved are described in detail in chapter 3.