An Integrated Job Assignment and Multiple Traveling Salesman Problem with Capacity and Delivery-Time Constraints for Land Logistics Providers in the International Supply Chain

Cross-border land logistics in international supply chains presents a complex decision-making challenge, as a logistics provider must simultaneously allocate jobs to vehicles and determine efficient delivery routes under strict capacity constraints and customer time windows. In this study, the linked job assignment and multiple traveling salesmen problem (mTSP) is formulated as a multiple objective optimization model that minimizes total travel distance, total travel time, and the number of late deliveries. To solve this problem, an adaptive goal-constrained genetic algorithm (AGCGA) is developed. First, a time-window-aware K-means clustering approach is used to generate feasible initial job assignments and estimate the required vehicle fleet. Next, a genetic algorithm with an embedded goal-programming fitness function is applied, where aspiration levels and penalty weights are adaptively adjusted during evolution to better explore the Pareto frontier. The proposed method is evaluated using a real-world-inspired dataset of 34 cross-border delivery locations between Thailand and Malaysia, with distance and travel-time matrices derived from Google Maps via web platforms. The results show that AGCGA outperforms NSGA-II and a greedy heuristic in terms of hypervolume, goal achievement, and computational stability. Overall, the framework provides a robust decision-support tool for balancing cost, service time, and delivery punctuality in international logistics.