Fairness in optimizing bus-crew scheduling process

Area of Science:

• Operations Research
• Transportation Science
• Artificial Intelligence

Background:

• Crew scheduling is complex, with fairness in working and idle times being a critical, yet often overlooked, factor.
• Existing models may not adequately address the nuanced cost structures and fairness constraints inherent in bus driver scheduling.

Purpose of the Study:

• To develop and validate a novel model for bus driver crew scheduling (CSP-BD) that explicitly incorporates fairness considerations.
• To enhance the hybrid ant-colony optimization (HACO) algorithm for improved efficiency and effectiveness in solving CSP-BD.

Main Methods:

• Proposing a model for crew scheduling for bus drivers (CSP-BD) with special cost structures and fairness constraints.
• Developing a hybrid ant-colony optimization (HACO) algorithm, enhanced with a Gamma heuristic function and specific selection rules.
• Analyzing cost relationships and algorithm performance using data from ten Beijing Public Transport Holdings bus lines.

Main Results:

• Demonstrated a valid approach for CSP-BD considering fairness in working and idle times.
• Identified indirect relationships between unfair cost and common, fixed, and extra costs.
• Showcased the HACO algorithm's efficiency, with tested bus lines completed in under 30 minutes.

Conclusions:

• The proposed HACO-based algorithm is a feasible and efficient optimization technique for bus driver crew scheduling.
• The model effectively addresses fairness constraints, offering a practical solution for large-scale transportation operations.
• Fairness considerations can be integrated into crew scheduling without compromising overall efficiency.