This presentation discusses a new program designed to investigate the effectiveness with which adaptive learning network (ALNJ analysis can be combined with linear array, phase steered, ultrasonic imaging techniques to provide an enhanced means for automatic data interpretations. The DARPA-sponsored program is being performed as a team effort between Adaptronics, Inc. and Battelle-Northwest. Battelle, under a subcontract from Adaptronics, is adapting the linear array imaging system being developed for the Electric Power Research Institute of Palo Alto, California, for use on this project. A special ultrasonic array will be developed to operate with the high-speed imaging system to acquire and record both specular and nonspecular signal information in both the time and frequency domains. Signal information from a multitude of simple and complex reflectors and defects wilI be recorded on the PDP 11 disk pack incorporated into the ultrasonic imaging system. Adaptronics will utilize the time·domain and frequency spectral data recorded from several thousand data points to develop algorithms and train networks which may describe uniquely the pattern of the reflections. The objective of the program is to provide a high-speed and automatic means for detecting, locating, sizing and displaying flaws in solid materials.