Prediction in business aviation using a data-driven model

Abstract

Nowadays, aviation companies can anticipate maintenance needs and ensure the availability (in advance) of spare parts for maintaining aircraft airworthiness and to avoid facing an unexpected AOG (Aircraft on Ground) status. The research paper explores supportive data-driven tools for predictive analytics across the civil aviation industry, focusing on how these technologies can transform line maintenance and spare parts inquiry forecasting. This study focuses on Business Aviation, where operational continuity is highly sensitive to delays due to the need for fast, flexible maintenance support. A structural framework was developed to describe and compare strengths, weaknesses and strategic priorities of relevant data-driven implementation. The priorities on the data-driven maintenance approach were explained in recent studies (Hang et al., 2023). Recent developments in predictive analytics allow component failure likelihoods to be estimated using historical removal records and aircraft utilization data. However, although failure probabilities can be predicted, the operational decision problem remains unresolved: how spare parts should be stocked, replenished, and allocated over time to minimize AOG risk while controlling logistics and operational costs. The research problem lies at the intersection of aircraft airworthiness management, maintenance planning, and logistics optimization. The challenge is characterized by uncertainty in component failures, limited inventory capacity, long replenishment lead times, quality of ordered spare part and the high penalty associated with delayed maintenance actions.