Seasonal and predictive analysis of transport fleet availability and reliability using long-term operational data

Abstract

This study introduces a novel empirical approach to analyzing seasonal variations in the availability and reliability of a transport vehicle fleet. While previous research has examined fleet reliability, few studies have integrated long-term operational data with complementary technical indicators and statistical modeling of seasonality. Using three key metrics - Fleet Availability Rate (FAR), Mean Time Between Failures (MTBF), and Mean Time to Repair (MTTR) - data from ten distribution vehicles operating over three years (2022–2024) were analyzed to identify recurring seasonal patterns. A linear regression model with seasonal dummy variables was applied to quantify the impact of weather conditions and operational intensity on vehicle availability. The results reveal a clear seasonal cycle: the lowest availability and highest failure rates occur between February and May, while summer and early autumn show near-optimal performance. The model demonstrated statistically significant differences between quarters and indicated a gradual long-term improvement in FAR. This study introduces a novel analytical and predictive framework that combines three reliability indicators with long-term operational data and regression-based seasonal modeling. The approach enables not only the identification of seasonal effects but also the prediction of fleet availability trends to support data-driven maintenance planning. These findings support more accurate forecasting of fleet availability and provide actionable guidance for optimizing maintenance schedules, resource allocation, and downtime risk management in transport operations. Overall, the results demonstrate how integrating operational data with seasonal regression models can improve predictive decision-making and optimize transport fleet reliability.