Absenteeism of Public Transport Drivers: Impact Analysis of Driving Shifts

In the public transport market, the absence of a driver, for unanticipated reasons, increases the cost and complexity of service management. To limit the adverse effects of absenteeism, it is essential to understand its causes and analyze possible operational levers to reduce the impact on the organization. This paper investigates the relationship between the characteristics of the work shifts assigned to drivers and their absenteeism. This study proposes an analysis methodology based on different data analysis techniques, applicable to structured data commonly collected in public transport companies. The methods presented were applied to a real-world scenario for demonstration purposes, thanks to data provided by the LPT (Local Public Transport) company Autoguidovie S.p.A. The methodological data analysis revealed a positive correlation between absences and breaks scheduled for drivers during consecutive runs. A weak, but still significant, correlation was also identified between absences and the end of the shift. Some driving shifts, clustered through a K-means clustering algorithm, revealed a positive correlation between absenteeism and out-of-residence. The machine learning model used to analyze the importance of the different characteristics of the driving shifts recognized the importance of the start and end times of the driving shifts and estimated the impact of monetary aspects such as compensation on the estimation of absences. Results do not presume to be conclusive, given the complexity of the phenomenon of absenteeism. This study aims to highlight the possibility of developing the analysis of this phenomenon, bringing attention to aspects that have been little considered until now.