Abstract
In the realm of personnel scheduling, where demand is subject to variability, the creation of a work schedule for irregular worker presents a significant challenge. This difficulty is amplified when the specific shift types are not predefined and when workers possess varying levels of availability and distinct skill sets. Additionally, due to legal and cost constraints, personal preferences, and fairness considerations, the scheduling problem becomes even more complex. The typical personnel scheduling process generally consists of two main stages: shift design and shift assignment. During the shift design stage, a scheduler designs a set of shifts and calculates the required number of workers to meet the demand. The shift assignment stage involves allocating workers to each individual shift, while satisfying various constraints to achieve certain objectives.The shift design and assignment stages can be approached as a single problem or solved as two separate problems. In this paper, a two-phase method is investigated, where the first phase is addressed by an integer programming approach, and the second phase is solved using a memetic approach. Real-world retail store scheduling instances are used to test and compare these methods. The solution time and quality suggest that the proposed methods can be effectively applied to solve the irregular scheduling problem commonly encountered in retail stores.