Conference Agenda

The Vehicle Scheduling Problem (VSP) represents a pivotal focus within transportation logistics. This study introduces novel practical objectives tailored explicitly for the context of intercity bus operations. The primary objective for VSPs focuses on minimizing fixed costs by minimizing the number of vehicles needed to serve all trips from the timetable. Additionally, the optimization targets reducing operational costs by strategically minimizing deadhead trips and waiting times.

This study introduces innovative metrics pertinent to the complexities of large-scale bus companies, which often rely on subcontractors for service execution. One such objective involves minimizing the number of distinct lines or trips in schedules, particularly crucial in cases where complete lines are entrusted to subcontractors. This strategic approach fosters schedule homogeneity among subcontractors, streamlining operational logistics and simplifying contingency planning in the event of breakdowns.

By preemptively optimizing schedules for subcontractors, bus companies can bolster their negotiating leverage and forge more advantageous partnerships. This paper presents a mixed-integer formulation addressing this multi-objective problem, complemented by a computational analysis of real-world instances. The findings underscore the efficacy of the proposed approach in enhancing operational efficiency and strengthening strategic positioning in intercity bus operations.

For environmental reasons, bus operators are planning to replace diesel buses with battery-electric ones in public transportation. However, scheduling electric buses is more complicated because of their limited battery capacity and long recharging times. Additionally, they require special charging infrastructure, which generally makes them less flexible than diesel buses.

In this study we address the single- and the multi-depot electric vehicle scheduling problem, in which a set of timetabled bus trips, each starting and ending at particular locations and times, should be carried out by a homogeneous fleet of electric buses. The buses start at their respective depot and return there after carrying out the service trips. Full or partial recharging is allowed either at charging stations located at the depot or at bus stops at the end of the respective service trips.

For the single- and multi-depot electric bus scheduling problem we present a three-index mixed-integer linear program where, in order to support the technology decision, we minimize lexicographically the number of vehicles to cover all timetabled trips, followed by the number of charging events during a day, and finally the energy spent on trips without passengers, so called deadhead trips. We reformulate the 3-index formulations into 2-index formulations and separate constraints of exponential size in a cutting plane fashion.