Advanced materials can provide specialized properties, or combinations of properties, that cannot be obtained in conventional materials. However, advanced materials generally require unusual processing operations in order to achieve their unique microstructures and the resulting greatly improved properties. These materials also tend to be expensive because of the high value added by unusual processing operations that may be labor intensive. Because the relationships among the processing parameters, microstructure, and resulting material properties and performance are not fully understood, and because the microstructure is difficult to control, reproducibility in these materials is often unsatisfactory. A very promising direction toward overcoming these difficulties involves intelligent processing of materials, a computer-based approach to automatically controlling the evolution of microstructure during processing.

In a conventional automated materials processing system, automation involves utilizing sensors to monitor process variables such as temperature and pressure. Data from these sensors are compared with preset values automatically in order to maintain these values with control devices. Nevertheless, the microstructure and properties often experience significant variations. The variations are detected after the fact either by destructive analysis in a quality control laboratory or by nondestructive evaluation (NDE) at the end of the manufacturing process.

In contrast, an intelligent processing system utilizes a new class of NDE sensors to characterize the microstructure of the material in real time. Moreover, the real-time data plus data from conventional process variable sensors are transmitted to a computerized decision maker.