Multi-store collaborative delivery optimization based on Top-K order-split.

Area of Science:

• Operations Research
• Supply Chain Management
• E-commerce Logistics

Background:

• Traditional retail order fulfillment often overlooks physical stores as key logistical hubs.
• The rise of new retailing necessitates models integrating online orders with in-store fulfillment capabilities.
• Existing research inadequately addresses joint optimization of order splitting and delivery from physical stores.

Purpose of the Study:

• To address the gap in optimizing online order fulfillment for traditional retailers using physical stores.
• To introduce the Multi-Store Collaborative Delivery Optimization (MCDO) problem.
• To develop an efficient algorithm for minimizing order fulfillment costs in a multi-store environment.

Main Methods:

• Development of the Multi-Store Collaborative Delivery Optimization (MCDO) problem formulation.
• Integration of Top-K breadth-first search and local search into a hybrid heuristic algorithm (TKILS).
• Optimization of search efficiency by controlling sub-orders and improving initial solutions with a greedy cost function.

Main Results:

• The proposed TKILS algorithm effectively optimizes both order splitting and delivery routing.
• Experiments demonstrate the algorithm's ability to minimize overall order fulfillment costs.
• Validation of the algorithm's effectiveness and applicability on synthetic and real-world datasets.

Conclusions:

• The MCDO problem and TKILS algorithm offer a novel solution for retailers leveraging physical stores for online order fulfillment.
• The hybrid heuristic approach significantly enhances the efficiency and effectiveness of order fulfillment processes.
• This research provides a practical and scalable method for optimizing multi-store collaborative deliveries in the new retail landscape.