Proceedings of the 5th Annual World Conference

of the Society for Industrial and Systems Engineering,

San Francisco, CA, USA
October 13-14, 2016

The Effect of Picking Congestion in Manual Order Picking Environments

M. Bataineh and M. Khasawneh

Systems Science & Industrial Engineering Department

State University of New York at Binghamton
Binghamton, NY, USA

Corresponding author's Email: mbatain1@binghamton.edu

Abstract: Manual order picking is considered as one of the costliest and time-consuming processes in a manual picker-to-order
system. The process starts with customer orders release. After distributing the orders to multiple order pickers, order picking
associates starts to fulfill customer demands. Picking is the process of collecting (picking) the order items from their storage
locations. Finding the most efficient route to visit all items locations is the main goal of optimizing order picking routing path.
Due to the fact that each warehouse runs its operations with multiple pickers, and the pickers may work in the same zone to
pick items on customer orders, congestion may occur during the picking time. This can affect the order picking finishing time.
The optimization solution obtained without considering the order picking congestion might not provide the most satisfaction
result. Most of the studies focus on the single picker to order systems, where the effect of the congestion does not exist.
However, in reality, multiple pickers are used. Therefore, this research will focus on the congested order picking problem, and
solve for the best picking route. Literature will be reviewed to identify the studies that consider the congestion effect in multiple
picker systems, and to review the models used to represent this problem.

Keywords: Manual Order Picking, Picker-To-Order

1. Introduction

Warehouses are buildings or controlled spaces which are responsible for storing different types of goods with different
sizes. Goods which are stored inside warehouses are usually received and placed in pre-designated storage locations (Muppani
& Adil, 2008). After orders from customers are received, these goods are retrieved from their locations and then prepared for
shipping. Therefore, warehouses play an important role in the middle of the supply chain, in which they connect manufacturers
with customers. Processes inside the warehouse are to control the inside operations, and those outside its borders are to control
its relations with other supply chain parties. Warehouses can be operated independently or within a manufacturer or retailer
site. They are usually built near highways, railroads, seaports, or airports for easy and efficient shipping. Warehouse buildings
usually have an inbound dock for receiving and an outbound dock for shipping; however, in other cases, receiving and shipping
can be completed through one dock.
Warehouses usually have different types of resources that manage and support different internal and external warehouse
operations. People, vehicles and trailers, and operating and management systems are considered the main resources of any
warehouse. Material processes in a warehouse handling can be differentiated based on the layout design and/or product
specifications. After products arrive at the inbound dock, they are received. Then, they are transferred to and placed at the
storage locations. Storage locations may be fixed (dedicated) for fixed types of products, or dynamically changed based on the
policies followed for each warehouse. The process where products are sent to the storage locations is also called put away.
During the receiving process, products can be checked for the type and condition, and counted. Customer orders initiate the
order picking process in which the ordered products are retrieved from their storage locations and then prepared for shipping.
As part of the supply chain, warehouse operations become more important as the logistics industry expands (higher volume
and product variety), and requires more accurate decisions within a limited time window. Among the logistic costs, 20% of the
total operational costs is due to warehouse operations (De Koster et al., 2007). This will lead to the need of more improvement
of warehouse operations.
As a building that is used to store different kinds of goods, a warehouse becomes a main component of the trading series,
used by different industrial companies, transporters, wholesale businesses, customers, etc. Warehouses usually occupy a large
space that can accommodate all stored items differentiated by storage needs. Warehouse docks are used for loading and
unloading. Modern warehouses, or sometimes known as distribution centers, are built and designed based on many optimization
studies. In addition, management systems become significant for processes in the warehouse. As they are becoming complex
(Karsek, 2013), they are now becoming more difficult to use. As a result, different software was developed with different

ISBN: 97819384960-8-0 256

Proceedings of the 5th Annual World Conference
of the Society for Industrial and Systems Engineering,
San Francisco, CA, USA
October 13-14, 2016
possibilities to meet different warehouse requirements; however, it is not easy to choose any software that can handle dynamic
changes in today’s warehouses (changes can be affected by many aspects in the warehouse).
A warehouse plays a major role in the supply chain. They mainly connect production and distribution processes, and
manage the inbound and outbound processes of receiving, storage, picking, and shipping which depends on characteristics of
each item. Warehouse definition expanded through the last few years. It is now no longer just for “storage”. Cost-conscious
manufacturers consider it as in the core of planning. This development is derived from the desire to optimize (reduce)
warehouse costs. Building new warehouses are based on many optimization studies for the different processes in the warehouse
and even for process redesigns. Warehouse optimization can be grouped into three categories (De Koster et al., 2007): 1.
Optimization of the technical structure, 2. Optimization of warehouse operations, and 3. Coordination between different
warehouse operations. This paper focuses on one of the operational processes inside the warehouse.
One of the main operational processes in the manual picker-to-order warehouse system is the order picking process. It
consists of receiving/reading orders by each picker, traveling to the location of each item on the order, picking, and then
delivering collected items to the shipping area. It also includes all non-value added activities such as waiting until the route is
clear and item searching. As a result, order picking can account for up to 55% of the total warehouse operating expenses
(Karsek, 2013). Therefore, traveling routes, restrictions, and policies can affect the total processing time, which impact the total
throughput. Improvement inside any warehouse could be related to the physical components of the warehouse such as the
building (layout) itself, or related to any other resources such as the people or vehicles. Other opportunities for improvements
could be related to the operational processes that take place inside the warehouse, and the management systems that are in
charge of all the entities of any warehouse.
Operational improvement has a large effect on the warehouse total processing time (and hence the total operating costs).
Manual order picking warehouses have multiple operational processes that control all the processes inside the warehouse.
Warehouse processes start with receiving goods at the inbound area. Receiving consists of checking the quantity and quality
of the goods. The received goods are sent to the specified storage locations, and put away by the operators. Goods are then
recollected based on the customer orders. In particular, the warehouse operators collect the specified quantities at each order,
and prepare the orders for shipping. Among these processes, order picking consumes most of the warehouse operational time
and cost. Furthermore, traveling to the storage locations consumes almost half of the picking time and cost (Around 50%
(Tompkins et al., 2003 & Horvat, 2012) to 55% (Bartholdi & Hackman, 2006)). For this reason, order picking can be considered
as a valuable area for improvement, and optimizing the traveling time can have a significant effect as part of this improvement
effort.
Order picking is the process of collecting items in customer orders (item retrieval) from their storage locations to prepare
for shipping. Order picking becomes an important process due to its large effect on the overall performance. There are two
main types of systems: picker-to-order, where pickers travel and collect items from their storage locations, and order-to-picker,
in which an automated retrieval system delivers the items to pickers (Henn et al., 2011).
Picking area in picker-to-order systems usually is the storage area, which consists of aisles that divide picking racks into
right and left picking shelves, or pallets (McDonald, 1995). Pickers ride different types of lift vehicles to collect items requested
in order lists. They start from a point, usually fixed for traveling and return, and through the picking aisles, stop at each location
of items to be picked, remove items from the order list, travel to drop-off area, and return to the starting point for a new order.
Some orders (due to their volumes or quantities) require more than one tour to be finished. Receiving customer orders is the
first step in this process. Each customer order includes the items required along with the quantity. Some warehouses divide the
storage locations into zones, where each designated picker is responsible to collect the items in their zones. All items are sorted
depending on the demand from each customer, and then shipped. In other warehouses, each picker can collect the full list of
the customer order items from all around the storage locations, and send the complete order to the shipping dock immediately.
Pickers travel inside the warehouse between the picking locations through picking aisles. Some warehouses define a fixed
travel direction in each aisle, and other warehouses allow the pickers to decide on the direction in each picking cycle.
One of the main resources in any warehouse is the number of pickers. With higher number of pickers, more orders can
be processed. On the other hand, congestion in the picking aisles may occur as the number of pickers increases. Congestion
can affect the order processing time which affects the efficiency of the overall picking process.
Routing in the picking (storage) area can be classified depending on the shape of the picking area. Rectangular
warehouses, which are the most common shape, use route restrictions, in which pickers can drive only with respect to these
policies. Some of the warehouses adopt simple routing policies, whereas others adopt more complex ones depending on the
conditions and operating systems in each warehouse. Five main strategies (Peterson, 1997) can be used in warehouses, varying
from simple to complex strategies. Each can be used depending on the warehouse shape (layout), order history characteristics,
item characteristics, etc. Transversal (S-shape), Return, Midpoint, Largest gap, and Composite policies.
Order picking improvement is driven by the combination of the customer expectations and the industry competition. In
the warehousing industry, each warehouse is constantly attempting to outperform other competitors to gain more market share.
At the same time customers become more demanding of a larger variety of products with higher expectations, which are quick

257
Proceedings of the 5th Annual World Conference
of the Society for Industrial and Systems Engineering,
San Francisco, CA, USA
October 13-14, 2016
orders fulfillment with a lower cost. As a result, this puts more pressure on the industry to lower warehouse operating costs and
reduce the fulfillment speed.
Warehouse improvement is in the form of faster and lower-cost operations. To remain competitive; warehouses must
receive, organize, pick, and ship products faster to keep them moving to the customers through the supply chain to make sure
that the customers receive their orders when they need them. It is estimated that around 55% of the total warehouse operating
costs are due to order picking (Karsek, 2013). As a result, order picking optimization is the area that most warehouses focus on
for improvements. If this operation becomes more efficient, the entire warehouse operations will improve. More efficient order
picking could result in cost reduction and decrease the time in which customers receive their orders. Most of the operating cost
associated with order picking process is labor cost, and most of the labor time is associated with traveling inside the warehouse
to collect the order items.
The problem of optimizing the order picking route can be considered as a traveling salesman problem, in which pickers
try to fulfill customer orders as soon as possible. One of the main restrictions inside any warehouse is the limited open space
for traveling. This can lead to more wasted time while the pickers are trying to finish any order. Traffic congestion has a
significant effect on the total finishing time as pickers will wait in many occasions to pass a picker (while picking) or wait to
get access to a place (occupied by another picker) to pick an item. The effect of congestion in the order picking problem will
be under investigation in this paper to study its impact on the actual warehouse environment.
In this paper, warehouse best picking route will be investigated based on rectangular warehouses with multiple storage
levels and manual order picking system. The best route will be designed based on the customer orders; i.e., the best route
changes as customer orders change. In each warehouse, route restrictions may apply to control the truck and people movements,
and to minimize the traffic and picking congestion. The main objective of this study is to design the optimum picking route
inside a warehouse. Optimum route is the route consists of minimum travel time and traffic congestion effect.

2. Literature Review

Meeting customer demand with low total warehouse costs is the main objective of order picking. Several constraints can
prevent the achievement of order picking objectives. The main constraints are related to the warehouse resources. Warehouse
operating costs and the form of constraints in optimization problems can be related to the number of each resource and their
roles (e.g., labor, layout, vehicles). Many researchers study this problem with different types of constraints.

2.1 Slotting Optimization

Different studies evaluate and compare different routing strategies. Liu (2004) develops a planning system, which
depends on the dynamic nature of the orders, and picking-area configuration. It deals with the storage layout and order picking
operation problems. This system consists of a formula for the enter-order-quantity rule, zero-one quadratic generalized
assignment model, and a heuristic procedure for stock location. This system efficiently solves the stock location problem in a
dynamic environment. To investigate the congestion effect on the operations in a multi-picker warehouse, a heuristic of the
storage assignment policy is studied, which minimizes the average time of order picking (Pan et al., 2012). Travel and waiting
times are considered to reduce the fulfillment time, and aisle number is determined to reduce blocking time and travel distance.
In another study (Brynzer, & Johansson, 1996), orders are grouped by the same characteristics, and a heuristic location
methodology is used to solve the storage location problem, which shortens the overall picking time.
There are three major decisions (Bozer, 1985) before slotting: 1. The quantity of stock keeping unit (SKU) to be stored
in the warehouse, 2. Replenishing time and sequence, and 3. The optimal location to store each SKU, which depends on both
the storage and access efficiency. Different techniques and methodologies are used to solve warehouse slotting optimization
problems. A knapsack-based heuristic is used (Bozer, 1985; Hackman et al., 1990), which involves the forward size as a
decision variable to minimize the costs (equipment and handling costs). A knapsack-based model is also proposed to minimize
the expected total labor time during busy time (Van den Berg et al., 1998). Branch-and-Bond is also used to minimize the
operation costs (Montulet et al., 1998). Also, simulated annealing and genetic algorithms are both used to solve a product-size
constraint problem, where items of the same product are stored near each other and the objective is to minimize the travel cost.

2.2 Routing Optimization

Many studies compare and evaluate different routing strategies. Several warehouse routing heuristics are evaluated
(Hwang et al., 2004) based on the cube per order index (COI). COI can be defined as the ratio of volume storage to inventory
for the SKU to the average number of occurrences of the SKU in order picking list for a given period of time (Haskett, 1963).

258
Proceedings of the 5th Annual World Conference
of the Society for Industrial and Systems Engineering,
San Francisco, CA, USA
October 13-14, 2016
Hwang et al., (2004) find Midpoint strategy has an advantage over Return and Traversal strategies. Caron et al. (1998) compare
and evaluate multiple routing strategies. They compare the S-heuristics (S-shape) with the Return heuristic in a low-level
manual picker-to-order system. This system implements a storage policy based on COI, in which cross aisles are increased
(Roodbergen and Koster, 2001). They evaluate different routing heuristics, and define a new method, aisle-by-aisle, in which
all the items are picked sequentially in each aisle, and the pickers can choose between traveling up to the end of each aisle or
coming back. S-shape, Combined (Composite), and Largest gap strategies are evaluated and compared with an optimal solution,
which is found by solving a traveling salesman problem with branch-and-bound.

2.3 Multiple Travel Salesman Problem

Multiple travel salesman problem (mTSP) is the general form of the well-known traveling salesman problem (TSP). In
mTSP, more than one salesman is considered. mTSP is related to vehicle routing problem (VRP), which calculates the minimum
number of vehicles to be used. Bektas (2006) defines the mTSP as the first stage in a procedure of a two-stage solution of a
VRP, in which cost in the second stage is calculated based on the results of the first stage.
Assignment-Based Formulation: Christofides et al. (1981) propose an assignment-based formulation by considering VRP
for a facility to meet the demand of different customers. Bektas (2006) summarizes the formulation of mTSP. The following
notation is suggested by Bektas (2006): the number of salesman is equal to m, with a set of nodes n, and a set of arcs A. C is
the cost matrix (c ij ), which is associated with A. Flow-Based Formulation: Bektas (2006) introduces the use of three index
formulation in mTSP, that was proposed by Christofides et al. (1981) in VRP. The decision variables and objective function
are as shown below. Different constraints can be added based on different warehouse restrictions and/or objectives.

3. Methodology for Warehouse Congestion

Congestion can occur when a warehouse order picking processes is interrupting. The interruption can be due to
overcrowding or clogging by another order picker. Congestion can be caused by different types of blocking (Huber, 2014):
1. Pick-Face Blocking: where two or more pickers are interested in picking from the same picking face.
2. In-the-Aisle Blocking: this type of blocking can occur in a narrow aisle. In this situation, the picker has to wait until
the aisle is free to pass. Aisle can be occupied by another picker which blocks the aisle in the front of other pickers
who need to pass.
3. In-the-Aisle Interface: this type can happen in a wider aisle. Pickers can pass each other. However, slower passing
velocity can affect the total picking time.
4. Total Aisle Blocking: the main cause of this type of blocking is when the aisle is reserved for only one picker. In this
situation, all other pickers have to wait outside until the aisle is clear (i.e. picker who reserves the aisle finished all
picks).
5. Cross Aisle Blocking: same as when the aisle is blocked by a picker and there is another one who wants to pass.
6. Depot Blocking: this situation can occur if two or more pickers try to use the depot at the same time.
In this paper, Pick-Face is the main type of blocking that will be discussed and evaluated. Picking congestion is the
term that will be used to refer the Pick-Face type. Picking congestion takes place when two or more pickers are trying to pick
from the exact same location, or pick from any other location above or below the occupied picking spot. This type of congestion
has the effect on the total picking time if pickers arrive at the blocking area at the same time. Due to congestion, the relation
between process throughput (number of orders processed per unit of time) and the number of pickers will not be increasing in
a linear relation, or the process time will be decreased. Therefore, as the number of pickers increases, the chance of congestion
will be increased. This will lead to increase of processing time due to congestion. Different papers study warehouse process,
and how to find the best route and best storage slot based. The effect of pick density and layout changes with respect to blocking
consideration was also studied. Many researchers focus on the warehouse processes considering congestion using simulation
software, analytical methods, stochastic processes, or queuing models (Huber, 2014).
Order pickers (or simply pickers) are the associates inside any warehouse. Their work is to fulfilled customer demands
by picking items required by customer orders. Customers set orders, these orders contain items to be picked and prepared for
shipping within a time limits. based on the warehouse picking strategy, pickers start to collect items from its storage locations
by traveling between these pick locations and retrieve the required quantities. Some warehouses divide their storage area into
zones, each is associated with one (or more) picker. This picker is only responsible to collect items from this zone. Other
warehouses just keep the entire storage area as one zone, and all pickers can pick from any storage location. In any case, pickers
goal is to fulfill customer orders within the time limits. Therefore, warehouse managers try to find the most efficient way to
determine the shortest path while traveling to pick. Pickers usually starts from a depot, in which picking orders are distributed.

259
Proceedings of the 5th Annual World Conference
of the Society for Industrial and Systems Engineering,
San Francisco, CA, USA
October 13-14, 2016
They start to travel to the first picking location. In some cases, traveling inside a picking aisle is restricted in one direction.
Pickers usually starts at the same time from the same point, and as they finish the first order they start with the next one.
Usually, same items can be found in different customer orders, which means that pickers will pick the same items while
fulfilling the orders. Picking the same item could cause the picking congestion if it is picked as the same time. Aisle congestion
could happen if one of the pickers is waiting behind another picker to pass. Congestion usually cause pickers to spend more
time waiting, and then more time to fulfill the orders. In some cases, the lowest distance route is not (necessarily) the fastest
one. So, some taking the longest route will save time more than just taking the shortest and spend extra time because of
congestion.
In this study, the warehouse will be in a rectangular shape, and the shelves will be used to store items that are ready for
picking, i.e. “replenishing area” is a different area. The picking equipment’s will have the ability to travel in both directions
and in a 3-D motion, in which it will be has the ability to travel in horizontal in vertical directions. Therefore, the travel distance
during any full order picking will be both the horizontal and vertical distances. A 3-D distance matrix for the distances between
the depot and all storage locations, and between the storage locations will be used to calculate the total traveled distance by all
pickers per full set of orders. The matrix will be transferred to 2-D matrix for the easy usage. The distances are from a real food
and paper warehouse, in which it will give more real results.
Solving an order picking problem is difficult. To impart more reality into the model, constraints are added in a way to
represent the actual restrictions inside the warehouse. As more constraints be annexed, more complexity be added to the model.
Model complexity could affect the solution efficiency, while adding more time to find the optimal (or best) solution, the
accuracy of the solution will be affected as well. Another factor could impact the solution efficiency, which is the size of the
problem. AS the problem get larger, finding the solution will take more time. For small size and less complexity problems,
commercial software is used to find an exact solution. As the problem get complicated, commercial software stuck or crash
while trying to find the solution, thus the exact solution cannot be find with this kind of problems. Heuristics methods are used
with such problem. Heuristics techniques are techniques used to help in finding an answer. These techniques are used to find a
solution, that is not guaranteed to be optimal, as well it will not tell how much it is close (more discussion will be presented in
the solution methodology section).
In order to reduce total picking time and increases the throughput, item storage locations or picking route are optimized
to find the most efficient way of storing or picking. In this model, the main goal is to find the optimal (or most optimal) picking
rout. Searching, picking, and set-up times will be assumed to be constant for the same number of picking orders so it will be
neglected (not calculated). On the other hand, this problem will be modeled as a multiple Salesman Problem (mTSP) with two
main objectives of finding the best picking route while including congestion effect. In order to find the shortest distance path
(by all pickers in the warehouse) and at the same time meet the demand and follow all the constraints, this problem will be
considered as multiple travel salesman problem (mTSP). Constraints in this problem is coming from a real warehouse operation.
MATLAB software will be used to solve this problem by coding the Tabu search (TS) algorithm to find the best solution for
this problem. The general procedure will be as follows:
• Randomly generate the orders (the customer orders will be generated, each with a multiple item to be picked from the
warehouse)
• Check all possible routing schemes for each order (Travel will start from the depot to all item’s “storage locations” then
come back to the depot)
• Check for congestion, if it has happened, apply penalty
• Compare all results
• Name the best so far sequence of items to be picked based on the minimum total distance of all pickers
Congestion penalty is applied as a waiting time to the objective function. Therefore, anytime congestion occur, an extra
distance will be added to the total traveled distance by each picker.

4. Preliminary Results, Conclusion, and Future Work

Two examples were used to address the congestion problem inside the warehouse. Each example with different number
of pickers, which means different number of orders. These orders will contain different number of items. In the first example,
example 1 is containing 10 items per order, while each order is containing 15 items per order in the second exam. These
examples are containing 10 and 15 items per customer orders in example 1 and 2 respectively. In both examples, total traveled
distance is calculated before and after addressing congestion (See Figure 1 & 2 for examples 1 & 2 respectively). The first step
is finding the best route using the coded Tabu search. After finding the best route, the algorithm will check for the congestion
and apply penalty based on the fact of existing congestion. In general, results show that as the number of pickers increases the
congestion occurrence percentage increases as a result. On the other hand, the percentage of congestion occurrence increases
as the number of orders items increases.

260
Proceedings of the 5th Annual World Conference
of the Society for Industrial and Systems Engineering,
San Francisco, CA, USA
October 13-14, 2016
The result shows that congestion has a high effect on the total picking time. Congestion could add more traveled time
(due to penalty) to the total picking time. Many papers did not address the congestion as a problem, and tried to find the best
picking route while minimizing (or canceling) the effect of multiple picker systems. Figure 3 shows how congestion effect
could be related to both, number of pickers working at the same time and the number of items in each (to be picked) order. The
results show that with 10 items in each order, total picking time could be increased between 9.6 to 10 % extra compared to
solutions without considering the congestion effect. As will, with 15 items the increase will be between 10.4 to 13.1 % to the
total traveled distance inside the warehouse.
As a conclusion, this paper addressed the congestion effect on the total traveling time inside a manual picker-to-order
warehouse system. Also, the model succeeded to include the congestion as a direct factor while finding the best picking route
inside the warehouse. As a future work, aisle congestion will be added and addressed along with the picking congestion as a
main factor that affect the picking travelling time. Big size problems will be addressed and solved using Tabu search and other
different heuristics for the sake of comparison.

Figure 1. Picking Congestion with 10 Items in Each Order

Figure 2. Picking Congestion with 15 Items in Each Order

261
Proceedings of the 5th Annual World Conference
of the Society for Industrial and Systems Engineering,
San Francisco, CA, USA
October 13-14, 2016

Figure 3. Percentage of Extra Traveled Time Due to Congestion Effect

6. References

Bektas, T. (2006). The multiple traveling salesman problem: an overview of formulations and solution procedures. Omega,
34 (3), 209-219.
Bozer, Y. A. (1985). Optimizing throughput performance in designing order picking systems.
Brynzer, H., & Johansson, M. (1996). Storage location assignment: Using the product structure to reduce order picking times.
International Journal of Production Economic, 46, 595-603.
Caron, F., Marchet, G. & Perego, A. (1998). Routing policies and COI-based storage policies in picker-to-part systems.
International Journal of Production Research, 36(3), 713-732.
Christofides, N., Mingozzi, A., & Toth, P. (1981). Exact algorithms for the vehicle routing problem, based on spanning tree
and shortest path relaxations. Mathematical programming, 20(1), 255-282.
De Koster, R., Le-Duc, T., & Roodbergen, K. J. (2007). Design and control of warehouse order picking: A literature review.
European Journal of Operational Research, 182 (2), 481-501.
Hackman, S. T., Rosenblatt, M. J., & Olin, J. M. (1990). Allocating items to an automated storage and retrieval system. IIE
transactions, 22(1), 7-14.
Henn, S., Koch, S., Gerking, H., & Wascher, G. (2011). A U-Shaped Layout for a Manual Order Picking System (Working
Paper No. 04/2011). Retrieved from Faculty of Economics and Management.
Horvat, M. (n.d.) An Approach to Order Picking Optimization in Warehouses.
Huber, C. (2014). Throughput Analysis of Manual Order Picking Systems with Congestion Consideration. KIT Scientific
Publishing.
Hwang*, H., Oh, Y. H., & Lee, Y. K. (2004). An evaluation of routing policies for order-picking operations in low-level
picker-to-part system. International Journal of Production Research, 42(18), 3873-3889.
Karásek, J. (2013). An Overview of Warehouse Optimization. International Journal of Advances in Telecommunications,
Electrotechnics, Signals and Systems, 2(3), 111-117.
Liu, C. (2004). Optimal Storage Layout And Order Picking For Warehousing. International Journal of Operations Research,
1(1), 37-46.
McDonald, S. C., Hertz, E. J., Smith, J. A., & Toto, G. (1995). U.S. Patent No. 5,468,110. Washington, DC: U.S. Patent and
Trademark Office.
Montulet, P., Langevin, A., & Riopel, D. (1998). Minimizing the peak load: an alternate objective for dedicated storage
policies. International journal of production research, 36(5), 1369-1385.
Muppani, V. R., & Adil, G. K. (2008). A branch and bound algorithm for class based storage location assignment. European
Journal of Operational Research, 189(2), 492-507.
Pan, J., Shih, P., & Wu, M. (2012). Storage assignment problem with travel distance and blocking considerations for a
picker-to-part order picking system. Computers & Industrial Engineering, 62(2), 527-535.

262
Proceedings of the 5th Annual World Conference
of the Society for Industrial and Systems Engineering,
San Francisco, CA, USA
October 13-14, 2016
Roodbergen, K. J. & De Koster, R. (2001). Routing order pickers in a warehouse with a middle aisle. European Journal of
Operational Research, 133(1), 32-43.
Van den Berg, J. P., Sharp, G. P., Gademann, A. J. R. M., & Pochet, Y. (1998). Forward-reserve allocation in a warehouse
with unit-load replenishments. European Journal of Operational Research, 111(1), 98-113.

263