ALL THE MAJOR MANUFACTURERS AUTOMATED THEIR FACILITIES AND ESTABLISHED COMPLEX COMPUTER SYSTEMS TO SUPPORT THEIR PERSONNEL IN THE 1980S TO COMPETE IN THE WORLD MARKETS. IN THIS DECADE, COMPETITION WILL BE EVEN MORE DEMANDING. IN ADDITION TO HUMAN INTELLIGENCE, COMPUTATIONAL POWER OF COMPUTERS, AND EXCELLENT REPEATABILITY OF ELECTROMECHANICAL SYSTEMS, THE SELF-LEARNING CAPABILITY OF THE MANUFACTURING EQUIPMENT WILL BE NEEDED. BACK PROPAGATION NEURAL NETWORKS ARE INTRODUCED AND THREE APPLICATIONS ARE DISCUSSED. THESE APPLICATIONS ARE MODELING OF QUASI STATIC MACHINE TOOL ERROR, DETECTION OF TOOL BREAAKAGE IN MILLING OPERATIONS, AND PREDICTION OF CHATTER DEVELOPMENT IN TURNING.