Development of a Decision Support System for Maintenance Troubleshooting: A Case Study in Aviation Maintenance, Repair, and Overhaul

This work presents the development of a Decision Support System (DSS) for prescribing corrective maintenance actions, a process known as troubleshooting, within the context of aviation Maintenance, Repair, and Overhaul (MRO). Framed within the methodological area of "Decision Making under Uncertainty," the primary goal of the DSS is to assist technicians and engineers in identifying the most probable root causes of failures and selecting the most appropriate maintenance actions to correct them, thereby reducing diagnostic time and increasing decision accuracy and consistency. The system integrates a hybrid approach combining rule-based decision logic, derived from the equipment's maintenance manual, with a probabilistic component built from historical maintenance records. The deterministic component maps symptoms and test results to documented root causes, ensuring traceability and compliance with technical guidelines. In parallel, the probabilistic component estimates the likelihood associated with each failure based on frequency observed in past cases, reinforcing the prioritization of recommended actions. Thus, the system does not predict the occurrence of failures; instead, it uses historical evidence to support the interpretation of past diagnostic results. The implementation is entirely developed in Python, utilizing data processing libraries, inference engines, and modular structures that facilitate continuous updates of both rules and historical data. The integration of these deterministic and probabilistic components creates a robust decision-support mechanism for scenarios where uncertainty and failure variability hinder a quick and accurate assessment. This work contributes to the advancement of intelligent tools applied to industrial maintenance, particularly in aviation, demonstrating the relevance of DSSs that combine technical information with probabilistic inference to improve the efficiency and effectiveness of troubleshooting operations.

Keywords: Decision Support System, Maintenance, Aviation, Troubleshooting, Uncertainty Management.