A Fuzzy Crew Rostering Model Based on Crew Preferences and Seniorities considering Training Courses: A Robust...

Area of Science:

• Operations Research
• Aviation Management
• Computer Science

Background:

• The crew scheduling problem is a critical aspect of airline operations, divided into crew pairing and crew rostering.
• Existing models often struggle to balance crew preferences, operational constraints, and uncertainties.

Purpose of the Study:

• To develop an advanced crew rostering model that maximizes crew satisfaction and adheres to regulations.
• To address the challenge of inconsistent crew assignments and uncertainty in seniority weights and time away from base.

Main Methods:

• Formulation of a rostering model incorporating crew preferences, seniority, regulations, and training attendance.
• Introduction of a novel scoring mechanism for desirable and undesirable assignments.
• Application of fuzzy logic for seniority weights and development of a robust counterpart for time away from base uncertainty.
• Utilization of a genetic algorithm (GA) for solving the crew rostering problem (CRP).

Main Results:

• The proposed model effectively assigns crew to pairings, maximizing weighted preferences and minimizing penalties.
• The genetic algorithm achieved an average optimality gap of only 0.5%, demonstrating high efficiency.
• Real-world data from Air India Airline showed a 61.59% reduction in undesirable rosters compared to previous methods.

Conclusions:

• The developed crew rostering model offers a significant improvement in crew satisfaction and operational efficiency.
• The novel scoring mechanism and handling of uncertainties provide a more robust and effective solution for airline crew scheduling.
• This research presents a pioneering approach to inconsistent crew rostering and demonstrates substantial reductions in undesirable outcomes.