To maintain a high level of safety, performance and reliability in controlled processes system errors, component faults and abnormal system operation must be detected promptly. Faults in process equipment can result in off-specification production, increased operating costs, the chance of line shut-down, and the possibility of detrimental environmental impact. Furthermore, prompt detection and diagnosis of process malfunctions are strategically important due to economic and environmental demands required for companies to remain competitive in world markets. Uncertainty of the true behaviour of the process being controlled, unknown disturbances and imprecise (but not necessarily faulty) measurements, all combine to make the task of early fault detection rather imprecise and difficult to achieve reliably. The more knowledge we can use about the process - whether this is in mathematical form, or descriptive (qualitative) or symbolic form, the better will be our chance of achieving reliable and robust fault detection. The principle of quantitative modeling for fault detection and isolation (FDI) is to compare actual and anticipated system responses, generated using mathematical models. The comparison between actual and expected plant behaviour can also be achieved using qualitative reasoning, qualitative modeling, fuzzy rule sets and methods from artificial intelligence (AI). This rather complex requirement to use all available system information has provided the motivation for the INCO-Copernicus project: Integration of qualitative and quantitative methods for fault diagnosis within the framework of industrial application (IQ2FD. The project has created the conditions for the proper development and integration of quantitative and qualitative methods of fault diagnosis by bringing together 11 well known research teams.