Order picking system design: the storage assignment and travel distance estimation (SA&TDE) joint method

Of all the warehouse activities, order picking is one of the most time-consuming and expensive. In order to improve the task, several researches have pointed out the need to consider jointly the layout of the warehouse, the storage assignment strategy and the routing policy to reduce travelled distances and picking time. This paper presents the storage assignment and travel distance estimation (SA&TDE) joint method, a new approach useful to design and evaluate a manual picker-to-parts picking system, focusing on goods allocation and distances estimation. Starting from a set of picking orders received in a certain time range, this approach allows to evaluate the combinations of product codes assigned to storage locations, aisles, sections or warehouse areas and to assess the most relevant ones, for the best location and warehouse layout, with the aim of ensuring optimal picking routes, through the application of the multinomial probability distribution. A case study is developed as well, in order to clarify the concept that underlies the SA&TDE joint method, and to show the validity and the flexibility of the approach, through the calculation of the saving at different levels of detail.     

Enhancing performance in order picking processes by dynamic storage systems

Increasing productivity and reducing labour cost in order picking processes are two major concerns for most warehouse managers. Particularly picker-to-parts order picking methods lead to low productivity as order pickers spend much of their time travelling along the aisles. To enhance order picking process performance, an increasing number of warehouses adopt the concept of dynamic storage where only those products needed for the current order batch are dynamically stored in the pick area, thereby reducing travel time. Other products are stored in a reserve area. We analyse the stability condition for a dynamic storage system with online order arrivals and develop a mathematical model to derive the maximum throughput a DSS can achieve and the minimum number of worker hours needed to obtain this throughput, for order picking systems with a single pick station. We discuss two applications of dynamic storage in order picking systems with multiple pick stations in series. In combination with simulation modelling, we are able to demonstrate that dynamic storage can increase throughput and reduce labour cost significantly.     

Impact of required storage space on storage policy performance in a unit-load warehouse

The performance of a storage policy in a warehouse is usually evaluated on the basis of the average one-way travel distance/time needed to store/retrieve a load. Dividing the storage space into zones based on item turnover frequency can reduce the travel distance. However, for a given number of stored items, a larger number of storage zones also requires more storage space, because of reduced space sharing between the items, which increases travel time. This study considers the required space consumption by storage zoning in comparing the performance of random, full turnover-based and class-based storage policies for a unit-load warehouse operated by a forklift in single-command mode. A generalised travel distance model that considers the required space consumption is developed to compare the performance of these policies. Results show that the one-way travel distance of a random policy decreases with the increase in skewness of the demand curve. By considering the required space consumption, a class-based storage policy performs generally better than a full turnover-based policy. In addition, the optimal warehouse shape factor (ratio of warehouse width to depth) appears to decrease with the skewness of the demand curve. Warehouse managers are advised to adopt a wide-shallow warehouse layout when the item demands are approximately equal, whereas a narrow-deep layout is preferred when the demand curves are steep.     

Evaluating order throughput time with variable time window batching

Given increased pressure to provide short delivery times, minimising customer order throughput time is a very important objective in warehousing operations. There are many factors that may affect the performance of an order picking system, such as layout of the warehouse, the storage strategy, the routing policy, the zoning method and the batching policy. In this study, we propose a simple travel time model and analyse the effect of order batching on the expected customer order throughput time with variable time window batching. In addition, we discussed the impact of some parameters on the batch size and the expected customer order throughput time. In the practice of Wuhan Dong Hon Logistics co. Ltd, the storage rack in the warehouse has two levels: the low storage level and the high storage level. These two levels are visited by pickers and forklifts, respectively. The classification improves the performance of order picking. In this study, we build the model based on this practice.     

Using a multiple-GA method to solve the batch picking problem: considering travel distance and order due time

Warehousing involves all activities related to the movement of goods such as receiving, storage, order picking, accumulation, sorting and shipping within warehouses or distribution centres. Among these activities, order picking is the most costly process because its operations are labour-intensive and repetitive. In this paper, we propose a batch picking model that considers not only travel cost but also an earliness and tardiness penalty to fulfil the current complex and quick-response oriented environment. This model is solved using a multiple-GA method for generating optimal batch picking plans. The core of the multiple-GA method consists of the GA__BATCH and GA__TSP algorithms. The GA__BATCH algorithm finds the optimal batch picking plan by minimizing the sum of the travel cost and earliness and tardiness penalty. The GA__TSP algorithm searches for the most effective travel path for a batch by minimizing the travel distance. To exhibit the benefits of the proposed model a set of simulations and a sensitivity analysis are conducted using a number of datasets with different order characteristics and warehouse environments. The results from these experiments show that the proposed method outperforms benchmark models.     

A Stock Location Model for Dual Address Order Picking Systems

A model is formulated to evaluate alternative stock levels and item locations for dedicated storage warehouses where storage and retrieval transactions are interleaved to economize on travel of the order picking vehicle. The model can be used to measure the order picking and inventory cost effects of warehouse layouts representing alternatives to application of economic lot sizing and cube per order index storage assignment. A simple, sequential search heuristic is proposed for generating such alternative warehouse layouts and applied to a sample problem.     

Optimal order picker routing in a conventional warehouse with two blocks and arbitrary starting and ending points of a tour

This paper investigates manual order picking, where workers travel through the warehouse to retrieve requested items from shelves. To minimise the completion time of orders, researchers have developed various routing procedures that guide order pickers through the warehouse. The paper at hand contributes to this stream of research and proposes an optimal order picker routing policy for a conventional warehouse with two blocks and arbitrary starting and ending points of a tour. The procedure proposed in this paper extends an earlier work of Löffler et al. (2018. Picker routing in AGV-assisted order picking systems, Working Paper, DPO-01/2018, Deutsche Post Chair-Optimization of Distribution Networks, RWTH Aachen University, 2018) by applying the concepts of Ratliff and Rosenthal (1983. “Order-picking in a Rectangular Warehouse: a Solvable Case of the Traveling Salesman Problem.” Operations Research 31 (3): 507–521) and Roodbergen and de Koster (2001a. “Routing Order Pickers in a Warehouse with a Middle Aisle.” European Journal of Operational Research 133 (1): 32–43) that used graph theory and dynamic programming for finding an optimal picker route. We also propose a routing heuristic, denoted S*-shape, for conventional two-block warehouses with arbitrary starting and ending points of a tour. In computational experiments, we compare the average order picking tour length in a conventional warehouse with a single block to the case of a conventional warehouse with two blocks to assess the impact of the middle cross aisle on the performance of the warehouse. Furthermore, we evaluate the performance of the S*-shape heuristic by comparing it to the exact algorithm proposed in this study.     

Data mining-based algorithm for storage location assignment in a randomised warehouse

Data mining has long been applied in information extraction for a wide range of applications such as customer relationship management in marketing. In the retailing industry, this technique is used to extract the consumers buying behaviour when customers frequently purchase similar products together; in warehousing, it is also beneficial to store these correlated products nearby so as to reduce the order picking operating time and cost. In this paper, we present a data mining-based algorithm for storage location assignment of piece picking items in a randomised picker-to-parts warehouse by extracting and analysing the association relationships between different products in customer orders. The algorithm aims at minimising the total travel distances for both put-away and order picking operations. Extensive computational experiments based on synthetic data that simulates the operations of a computer and networking products spare parts warehouse in Hong Kong have been conducted to test the effectiveness and applicability of the proposed algorithm. Results show that our proposed algorithm is more efficient than the closest open location and purely dedicated storage allocation systems in minimising the total travel distances. The proposed storage allocation algorithm is further evaluated with experiments simulating larger scale warehouse operations. Similar results on the performance comparison among the three storage approaches are observed. It supports the proposed storage allocation algorithm and is applicable to improve the warehousing operation efficiency if items have strong association among each other.     

Routing methods for warehouses with multiple cross aisles

This paper considers routing and layout issues for parallel aisle warehouses. In such warehouses order pickers walk or drive along the aisles to pick products from storage. They can change aisles at a number of cross aisles. These cross aisles are usually located at the front and the back of the warehouse, but there can also be one or more cross aisles at positions in between. We describe a number of heuristics to determine order picking routes in a warehouse with two or more cross aisles. To analyse the performance of the heuristics, a branch-and-bound algorithm that generates shortest order picking routes is used. Performance comparisons between heuristics and the branch-and-bound algorithm are given for various warehouse layouts and order sizes. For the majority of the instances with more than two cross aisles, a newly developed heuristic appears to perform better than the existing heuristics. Furthermore, some consequences for layout are discussed. From the results it appears that the addition of cross aisles to the warehouse layout can decrease the handling time of the orders by lowering average travel times. However, adding a large number of cross aisles may increase average travel times because the space occupied by the cross aisles has to be traversed as well.     

Managing warehouse efficiency and worker discomfort through enhanced storage assignment decisions

Humans are at the heart of crucial processes in warehouses. Besides the common economic goal of minimising cycle times, we therefore add in this paper the human well-being goal of minimising workers’ discomfort in the context of order picking. We propose a methodology for identifying the most suitable storage location solutions with respect to both goals. The first step in our methodology is to build data-driven empirical models for estimating cycle times and workers’ discomfort. The second step of the methodology entails the use of these empirically grounded models to formulate a bi-objective assignment problem for assigning products to storage locations. The developed methodology is subsequently tested on two actual warehouses. The results of these practical tests show that clear trade-offs exist and that optimising only for discomfort can be costly in terms of cycle time. Based on the results, we provide practical guidelines for taking storage assignment decisions that simultaneously address discomfort and travel distance considerations.     

A correlated storage location assignment problem in a single-block-multi-aisles warehouse considering BOM information

This paper deals with a correlated storage location assignment problem by considering the production bill of material (BOM) information. Due to the large number of parts in a BOM, the picking capacity constraint is considered. A mathematical model is formulated and a multi-stage heuristic is proposed. The heuristic relaxes the close interrelationships within the original problem through an improvement algorithm and an iterated approach to incorporate the effect of BOM splitting. In order to evaluate the performance of the heuristic, numerical experimentation is conducted in a single-block-multi-aisles warehouse, by using a randomly generated data set.     

An integrated storage assignment method for manual order picking warehouses considering cost,workload and posture

This paper presents an integrated storage assignment method for low-level picker-to-parts order picking warehouses taking into account economic and ergonomic objectives. Three different pallet rack layouts are studied in this paper, namely (a) picking from full pallets on the floor, (b) picking from half-pallets on the floor, and (c) picking from half-pallets on the upper rank of the shelf. First, cost functions are developed to assess the total order picking performance impact of these different pallet rack layouts. Second, with regard to workload, the metabolic cost and energy expenditure rates for picking from the different rack layouts under study are derived. Third, for assessing the working posture during order picking, the Ovako Working Posture Analysing System index is used where the required data is collected using a motion capturing system. The developed models are combined to propose a heuristic storage assignment procedure that supports the decision of which item to store on which pallet. The developed storage assignment method is then applied to an industrial case study. The results of the paper support warehouse managers in assessing the order picking storage assignment from an ergonomics viewpoint and in estimating its impact on financial order picking performance.     

Enhancing the order picking process through a new storage assignment strategy in forward-reserve area

This paper reexamines the order picking process in a warehouse facing the challenges that e-commerce brings about and which are characterised by a very large number of small sized orders and returns. Implementing effective storage assignment strategies combined with efficient batching, in this context, is very fundamental to keep the warehouse’s responsiveness and order completion times up to the standards. This paper investigates a new storage assignment strategy, initially implemented by a large shoes and footwear wholesaler, to enhance the performance of the order picking process in its warehouses. The impact of this new storage assignment strategy on the performance of the order picking process is analysed via simulation. The performance of the system, measured in terms of total picking time as well as order lead time while taking congestion into account, is compared to the cases where conventional storage assignment strategies are implemented. A full factorial design is set up and the simulation output is statistically analysed. The results of this analysis are reported and thoroughly discussed. Attending to the results, the proposed strategy presents a remarkable potential to shorten total picking travel distances and order completion time and consequently customer satisfaction.     

Order picking in a parallel-aisle warehouse with turn penalties

In many real-life routing problems, incorporating the negative effects of turns is an important, but often overlooked aspect. This is especially true for order picking in warehouses, where making the turns not only decreases the picking efficiency by reducing the speed of the vehicle, but it also results in other unquantifiable effects such as vehicle tipovers, increased congestion and increased risk of collision with pedestrians or other vehicles. In this paper, we consider the order picking problem in a parallel-aisle warehouse by taking into account the number and effect of the turns. In particular, we show that the problem of minimising the number of turns, minimising travel time under turn penalties, the biobjective problem that involves turn and travel time minimisation as separate objectives, and the triobjective problem with U-turn minimisation as a third objective can all be solved in polynomial time. Our computational results show that the algorithms we develop can generate the corresponding Pareto front very quickly, and significantly outperform heuristic approaches used in practice.     

Heuristic routing methods in multiple-block warehouses with ultra-narrow aisles and access restriction

This paper focuses on multiple-block warehouses with ultra-narrow aisles and access restriction. These new features observed from one of the largest online retailers in China allow order pickers enter pick aisles from specific entrances but prohibit them from traversing the aisles. This impedes the application of traditional heuristic order picking methods. To address the order picking problem in such warehouses, we propose six heuristic routing methods by extending the basic Return, Largest Gap and Mid-point methods for the single-block warehouse. These six heuristic methods are named RNA, LNA, MNA, RNAP, LNAP and MNAP, respectively. The major improvements are achieved through setting rules with respect to determining the access mode of aisles as well as changing working aisles. Using real order information, a comprehensive simulation for comparison is conducted to evaluate the effectiveness of our improved routing methods under 12 warehouse layouts. The simulation results demonstrate that LNAP achieves the shortest average picking routes in most scenarios. The impacts of warehouse layout on performance measurements are analysed as well. It is ascertained that setting more cross aisles and connect aisles helps mitigate the negative impacts.     

A model for warehouse layout

This paper describes an approach to determine a layout for the order picking area in warehouses, so that the average travel distance for the order pickers is minimized. We give analytical formulas that can be used to calculate the average length of an order picking route under two different routing policies. The optimal layout can be determined by using these formulas as the objective function in a nonlinear programming model. The optimal number of aisles in an order picking area appears to depend strongly on the required storage space and the pick list size.     

基于荷兰式拍卖机制的RMFS货位指派研究

目的对移动机器人拣货系统(RMFS)的货位指派问题进行研究,提出基于荷兰式拍卖机制的货位指派模型来提高RMFS的拣货效率。方法构建荷兰式拍卖模型并按周转率对指定区域的SKU进行逐步调整,以货位指派匹配度为期望指标,实现SKU的需求模式与仓库存储结构之间的合理匹配。结果通过与随机指派模型对比发现,在不同仓库规模、需求偏度、订单规模等情况下,基于荷兰式拍卖的货位指派方法可以使拣货路程下降21.15%、工作时间平均下降20.57%左右。结论与传统指派方法相比,提出的货位指派模型可以大幅度降低RMFS系统的拣货距离和时间,大幅度提升在线零售企业的RMFS的拣货效率。     

基于鱼骨型仓库布局的多车拣选路径问题优化

为了推动鱼骨型仓库在实际场景下的应用,针对鱼骨型仓库布局下的拣货路径优化问题,构建待拣货点距离计算模型和以有载重、容积限制的多车拣货距离最短为总目标的拣选路径优化模型。考虑遗传算法(GA)全局搜索能力强、粒子群算法(GAPSO)收敛速度快以及蚁群算法(ACO)较强的局部寻优能力,提出一种解决拣选路径优化模型的混合算法(GA-PSO-ACO)。通过不同订单规模的仿真实验,得出该混合算法在适应度值、迭代次数、收敛速度等方面均优于GA算法和GAPSO算法,且在订单规模较大时,平均适应度值约降低8%,有效缩短了总拣选距离,验证了混合算法在解决鱼骨型仓库布局下的拣货路径问题的先进性和有效性,为解决此类仓库内部的拣货路径问题提供新的解决方法和思路。     

Optimal routing in an automated storage/retrieval system with dedicated storage

We address the sequencing of requests in an automated storage/retrieval system with dedicated storage. We consider the block sequencing approach, where a set of storage and retrieval requests is given beforehand and no new requests come in during operation. The objective for this static problem is to find a route of minimal total travel time in which all storage and retrieval requests may be performed. The problem of sequencing a list of retrievals is equivalent to the Traveling Salesman Problem (TSP), and thus NP-hard in general. We show that the special case of sequencing under the dedicated storage policy can be solved in polynomial time. The results apply to systems with arbitrary positions of the input and output stations. Tlus generalizes the models in the literature, where only combined input/output stations are considered. Furthermore we identify a single command area in the rack. At the end we evaluate the model against heuristic procedures.     

Clusters and filling-curve-based storage assignment in a circuit board assembly kitting area

Kitting is the process that gathers all of the components necessary to assemble a batch of circuit boards on production machines. An objective in kitting storage design is to minimize the travel involved in collecting the components in the storage area so as to decrease labor costs. From outward appearances, the kitting process is similar to order picking in a warehouse. Closer observation, however, reveals that its unique characteristics favor a cluster-based allocation over the storage strategies usually adopted in warehouses. We present a clustering and cluster assignment method. In the clustering method we develop a new objective function and incorporate it into a genetic algorithm. In the cluster assignment method we first develop a new index for cluster assignment priorities. We then prove optimum assignments of clusters under restrictive conditions and extend the result to realistic storage configurations using filling curves. We analyze the properties affecting the quality of filling curves and develop a class of filling curves with good performance characteristics. Finally, we perform numerical analyses to show that the cluster and filling-curve-based assignment in the kitting area can reduce travel distances.