With the rapid development of e-commerce in recent years, the delivery orders of parcels are characterized by small batches, multiple varieties, frequent orders, short delivery cycles, and high precision requirements, which requires more efficient sorting operations. Logistics companies have gradually begun to use “double-layer automatic sorting systems” to improve the sorting efficiency and shorten the sorting time. In order to further improve the sorting efficiency, the “group sorting” strategy is recently adopted by some companies to improve the pallet utilization. This strategy combines multiple parcels with the same destination according to certain rules, and then sorts the normalized and standardized large units. This paper studies an optimization problem related toal locating packages, pallets, sorting boxes, and supply stations in a double-layer sorting system considering the group sorting strategy.An integer programming model is established, an efficient solution algorithm based on column generation is developed, and acceleration techniques for pricing problem are implemented to further shorten the solution time. Experimental results validate the effectiveness of applying the group sorting strategy in a double-layer automatic sorting system.In addition, based on some sensitivity analysis experiments on the proposed model and algorithm, this paper proposes managerial implications for technology adoption and equipment setting in automatic sorting systems.