A System for Detecting Non-Conformities in Industrial Manufacturing Processes in the Automotive Sector Using Fuzzy Logic

Abstract

Increasing industrial competitiveness demands stable and continuously monitored production processes. This work presents a Non-Conformance Detection System based on Fuzzy Logic, applied to quality control in manufacturing. Using the PPM (Parts Per Million) indicator, the system evaluates process performance and identifies the risk level of non-conformities. Scenarios with PPM between 20 and 11,000 were analyzed, covering conditions from ideal to critical situations. The fuzzy model, composed of triangular and trapezoidal membership functions, expert rules, and centroid defuzzification, showed smooth transitions and greater sensitivity compared to deterministic methods. The results confirm its practical applicability in the automotive industry, assisting in risk identification, rapid decision-making, and continuous process improvement.