Dynamic maintenance scheduling approach under uncertainty: Comparison between reinforcement learning, genetic algorithm simheuristic, dispatching rules

Ruiz-Rodríguez, Marcelo Luis; Kubler, Sylvain; Robert, Jérémy; Le Traon, Yves

doi:10.1016/j.eswa.2024.123404

Cited by 3 publications

(2 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Referring to the aircraft maintenance rescheduling [44] applies a mathematical optimisation model for the efficient maintenance tasks rescheduling problem in a disruptive environment. Ruiz-Rodríguez et al [37] analyse the maintenance scheduling problem with a focus on the impact of uncertainty on the failure distribution and time to repair.…”

Section: Trackwork Reschedulingmentioning

confidence: 99%

“…Unexpected events and disruptions that can affect trackwork scheduling include equipment unavailability, staff shortages, changes in work priorities, and changes in work order. According to [37], uncertainty in maintenance is also linked to maintenance type, maintenance duration, technician availability, machine availability, failure distribution and joint schedule.…”

Section: Trackwork Reschedulingmentioning

confidence: 99%

See 1 more Smart Citation

Stability assessment of railway trackwork scheduling in Sweden

Ivina,

2024

Eur. Transp. Res. Rev.

View full text Add to dashboard Cite

Ensuring the reliability of railway transportation is heavily dependent on the quality of its infrastructure. In this regard, renewal and maintenance of the railway track infrastructure, referred to as trackwork, play a vital role. However, trackwork execution requires temporary capacity restrictions for train traffic. Therefore, harmonising the train and maintenance schedules is critical but challenging to accomplish when one is frequently changing. This paper explores and models the nature of trackwork schedule instability at the tactical level of the scheduling process. We analyse data from one year of trackwork rolling horizon plans, focusing on weekly changes at eight key trackwork locations across Sweden’s railway network. Our study considers various factors that may affect schedule stability, such as track type, location, time of day, train traffic intensity, and the type of prevailing traffic. We find that schedule instability increases as the rolling horizon plan approaches its end. The regression analysis reveals that the most significant predictors of changes in trackwork schedules include previous changes, track type (single vs. double), work location (at station vs. between stations), and the timing of trackwork (daytime vs. nighttime and month). These provide insights to trackwork planners in making informed and proactive decisions about trackwork timeslot allocation.

show abstract

Section: Trackwork Reschedulingmentioning

confidence: 99%

Section: Trackwork Reschedulingmentioning

confidence: 99%