GRASP applied to the Order Batch Separation Problem

Order picking is one of the most critical operations in warehouse logistics, particularly in the context of e-Commerce, where speed and accuracy are paramount. This paper addresses the Order Picking in Waves problem (OSW), where a subset of customer orders is grouped for simultaneous collection to optimize operational efficiency. The problem consists of selecting a set of orders and the aisles to be traversed so that the picking efficiency is maximized. In this case, the objective is to maximize the average number of items picked per aisle, while respecting capacity and product availability constraints. This problem is motivated by a real-life warehouse logistics problem and by a computational model that was proposed as a optimization challenge by Mercado Libre at the SBPO 2025. This work proposes a GRASP heuristic for this problem, that showed to be more efficient than the proposed baseline.