ARTIFICIAL INTELLIGENCE SUPPORT FOR STATISTICAL PROCESS CONTROL OF DISCRETE MANUFACTURING PROCESSES

THIS RESEARCH ADDRESSES THE PROBLEM OF HOW TO ACHIEVE HIGH QUALITY PRODUCTS AT MINIMUM COST. MANUFACTURERS UNDERSTAND THAT THE WAY TO ACHIEVE QUALITY IS BY CONTROLLING THE PROCESS SO AS TO MINIMIZE DEFECTS RATHER THAN BY REJECTING PARTS IN AN INSPECTION PROCESS. TO DO THIS, THEY NEED TO APPLY STATISTICAL PROCESS CONTROL PRINCIPLES TO THE MANUFACTURING PROCESS. PROPER APPLICATION OF STATISTICAL PROCESS CONTROL REQUIRES ALEVEL OF MATHEMATICAL SOPHISTICATION WHICH IS WAY BEYOND THEEDUCATIONAL LEVEL OF MOST PRODUCTION OPERATORS, TECHNICIANS,AND SUPERVISORS. MOST MANUFACTURERS CANNOT AFFORD TO EMPLOYPEOPLE WITH THE REQUIRED KNOWLEDGE. A POTENTIAL SOLUTION TO THIS DILEMMA IS TO PUT THE REQUIRED KNOWLEDGE INTO COMPUTER SYSTEMS WHICH ASSIST SHOP FLOOR PERSONNEL IN PERFORMING STATISTICAL PROCESS CONTROL. THE OBJECTIVE OF THE PROPOSED PROJECT IS TO EXTEND PRIOR RESEARCH INTO REAL-TIME ARTIFICIAL INTELLIGENCE SYSTEMS TO ENABLE ITS APPLICATION TO THIS PROBLEM AND TO TEST THE VIABILITY OF THIS SOLUTION IN ACTUAL MANUFACTURING OPERATIONS. PRIOR RESEARCH BY THE INVESTIGATORS HAS DEMONSTRATED THE VIABILITY OF USING ARTIFICIAL INTELLIGENCE SYSTEMS TO PROVIDE ADVISE IN REAL-TIME TO PILOTS. THIS TECHNOLOGY IS PROBABLY APPLICABLE TO THE QUALITY CONTROL PROBLEM PROVIDED THAT APPROPRIATE OPERATOR INTERFACES CAN BE DEVELOPED.