Proceedigs of the 15th IFAC Symposium on

Proceedigs
Information
Proceedigs of the 15th
Control IFAC Symposium
Problems on
in Manufacturing
Proceedigs ofof the
the 15th
15th IFAC
IFAC Symposium
Symposium on
on
Available online at www.sciencedirect.com
Information
May Control
11-13, 2015.
Information Problems
Ottawa,
Control in
in Manufacturing
Canada
Problems Manufacturing
Information Control Problems in Manufacturing
May
May 11-13,
11-13, 2015.
2015. Ottawa,
Ottawa, Canada
Canada
May 11-13, 2015. Ottawa, Canada
ScienceDirect
IFAC-PapersOnLine 48-3 (2015) 11521154
Warehouse Design under Class-Based Storage Policy
Warehouse
Warehouse Design
Design under
of Shuttle-Basedunder Class-Based
Class-Based
Storage Storage
Storage
and Retrieval Policy
Policy
System
of Shuttle-Based Storage and Retrieval System
of Shuttle-Based Storage and Retrieval System
B.Y. Ekren*, Z. Sari**, T. Lerher***
B.Y. Ekren*,
B.Y.
B.Y. Ekren*, Z.
Ekren*, Z. Sari**,
Z. Sari**, T.
Sari**, T. Lerher***
T. Lerher***
Lerher***



*Department of Industrial Engineering, Izmir University of Economics, Balcova, Izmir, 35330 Turkey
*Department
*Department of Industrial Engineering, Izmir University of
*Department of
of Industrial
Industrial Engineering,
Engineering,
(Tel: Izmir
Izmir
232-279-2525; University
University
e-mail: of Economics,
of Economics,
banu.ekren@ Balcova,
Balcova, Izmir,
Balcova,
ieu.edu.tr)
Economics, Izmir, 35330
Izmir, 35330 Turkey
35330 Turkey
Turkey
(Tel:
(Tel:
**Manufacturing Engineering Laboratory 232-279-2525;
232-279-2525;
of Tlemcen e-mail:
e-mail:
(MELT), banu.ekren@
banu.ekren@
University of ieu.edu.tr)
ieu.edu.tr)
Tlemcen,
(Tel: 232-279-2525; e-mail: banu.ekren@ ieu.edu.tr) Algeria (e-mail: z_sari@mail.univ-
**Manufacturing
**Manufacturing Engineering
**Manufacturing Engineering Laboratory
Engineering Laboratory of
Laboratory of Tlemcen
of Tlemcen (MELT),
Tlemcen (MELT), University
University of
University
tlemcen.dz)
(MELT), of Tlemcen,
of Tlemcen, Algeria
Tlemcen, Algeria (e-mail:
Algeria (e-mail: z_sari@mail.univ-
(e-mail: z_sari@mail.univ-
z_sari@mail.univ-
*** Faculty of Mechanical Engineering, University oftlemcen.dz)
tlemcen.dz)
Maribor,
tlemcen.dz) Smetanova 17, SI-2000, Maribor, Slovenia, (e-mail:
***
*** Faculty
*** Faculty of
Faculty of Mechanical
of Mechanical Engineering,
Mechanical Engineering, University
Engineering, University of
of Maribor,
Maribor, Smetanova
Smetanova 17,
tone.lerher@um.si)}
University of Maribor, Smetanova 17, SI-2000,
17, SI-2000, Maribor,
SI-2000, Maribor, Slovenia,
Maribor, Slovenia, (e-mail:
Slovenia, (e-mail:
(e-mail:
tone.lerher@um.si)}
tone.lerher@um.si)}
tone.lerher@um.si)}
Abstract: In this study, we aim to find out the best rack design for shuttle-based storage and retrieval
Abstract: In
In this
system (SBS/RS)
Abstract:
Abstract: In this study,
this under we
study,
study, we aim
aim toto find
class-based
we aim to find out
out the
storage
find out the best
policy
the best rack
rack design
best(CSP).
rack design
SBS/RS
design for
for is
for shuttle-based
shuttle-based
a new technology
shuttle-based storage
storage in
storage and retrieval
andautomated
and retrieval
retrieval
system
system
storage (SBS/RS)
(SBS/RS)
and under
under
retrieval class-based
class-based
system (AS/RS) storage
storage
which ispolicy
policy (CSP).
(CSP).
developed forSBS/RS
SBS/RS
high is
is a
a
transactionnew
new technology
technology
environments in
in automated
automated
where mini-
system (SBS/RS) under class-based storage policy (CSP). SBS/RS is a new technology in automated
storage
storage
load and
and
AS/RS retrieval
retrieval
crane maysystem
system
not be(AS/RS)
(AS/RS)
able to which
which
keep is
is
pace developed
developed
with the for
for high
high
transaction transaction
transaction
rate neededenvironments
environments
storage and retrieval system (AS/RS) which is developed for high transaction environments where mini- over a givenwhere
where
number mini-
mini-of
load
load AS/RS
AS/RS
storage crane
crane
locations. may
may
We not
not be
be
consider able
able to
to
several keep
keep
rack pace
pace with
with
design the
the
conceptstransaction
transaction
in terms rate
rate
of needed
needed
number
load AS/RS crane may not be able to keep pace with the transaction rate needed over a given number of over
over
of a
a given
given
aisles, number
number
tiers and of
of
bays
storage
storage locations.
locations. We
locations.
in the warehouse.
storage We
WeTheconsider
consider several
several
performance
consider severalof rack
thedesign
rack
rack design
system
design concepts
concepts in
in terms
in
is evaluated
concepts terms
termsin of
of
of number
number
terms
number of
of aisles,
of aisles, tiers
aisles,
of utilizations of and
tiers
tiers lifts bays
and
and bays
and
bays
in
in the
the warehouse.
warehouse.
storage/retrieval The
The
devices performance
performance
and cycle of
of
times the
the
of system
system is
is evaluated
evaluated
storage/retrieval in
in terms
terms
transactions. We
in the warehouse. The performance of the system is evaluated in terms of utilizations of lifts and of
of utilizations
utilizations
utilize of
of
simulation lifts
lifts
for and
and
the
storage/retrieval
storage/retrieval
modeling purpose. devices
devices
The and
and
resultscycle
cycle times
times
indicate of
of
that storage/retrieval
storage/retrieval
CSP works better transactions.
transactions.
under high We
We
rise utilize
utilize
warehouse
storage/retrieval devices and cycle times of storage/retrieval transactions. We utilize simulation for the simulation
simulation
design. for
for
So, the
the
this
modeling
may createpurpose.
modeling
modeling purpose. The
The results
less number
purpose. The of liftsindicate
results
results indicate that
that CSP
requirement
indicate that CSP works
works
in the
CSP better
better
system.
works under
underthe
Hence,
better under high
high
high rise
rise warehouse
warehouse
rise design design.
warehouse
warehouse design.
with CSP
design. So, this
So,may
So, this
this
may
may create
create
tend create
may less
less
to haveless number
number
lower of lifts
of lifts
investment
number of requirement
lifts cost.
requirement
By this in
requirement in
in the
study,
the system.
the system.
system. Hence,
Hence,
we also Hence,
aim to thethe
the warehouse
warehouse
provide design
designand
practitioners
warehouse design with CSP
withacademia
with CSP may
CSP may
maya
tend
tend to
to have
to have
significant
tend have lower
lower
insight
lowerforinvestment
investment cost.
cost. By
cost.
SBS/RS design
investment By this
this
under
By study,
study,
thisseveral we
we also
we
study, rack also aim
aim
design
also to
to provide
to provide
aimconcepts
providefor practitioners
practitioners
CSP.
practitioners and
and academia
and academia aaa
academia
significant
significant
significant insight
insight
insight for
for SBS/RS
SBS/RS
for SBS/RS design
design under
under
design under several
several rack
rack
several rack design
design concepts
concepts
design concepts for
for CSP.
CSP.
for CSP.
2015, IFAC
Keywords: (International
Automated Federation
storage of Automatic
and retrieval system,Control)
SBS/RS,Hosting by Elsevier
class-based Ltd.simulation,
storage, All rights reserved.
automated
Keywords:
Keywords:
warehouse. Automated
Automated storage
storage and
and retrieval
retrieval system,
system, SBS/RS,
SBS/RS, class-based
class-based storage,
storage,
Keywords: Automated storage and retrieval system, SBS/RS, class-based storage, simulation, automated simulation,
simulation, automated
automated
warehouse.
warehouse.
warehouse.

1. INTRODUCTION In this study, we consider an SBS/RS with two lifting
1. INTRODUCTION In
In this
In this study,
this
tables attached on leftwe
study,
study, we
we consider
consider
and an
an
right side
consider SBS/RS
SBS/RS
an of with
with
the lifting
SBS/RS two
two lifting
two lifting
withmechanism.
lifting
1. INTRODUCTION
1. INTRODUCTION
Increasing trend towards more product variety and tables
tables
Our attached
attached
aim is to on
on left
left
explore and
and
the right
right
best side
side
rack of
of the
the
design lifting
lifting
for mechanism.
mechanism.
SBS/RS under
tables attached on left and right side of the lifting mechanism.
Increasing
Increasing trend
trend towards
towards more product variety and Our
Our aim is to explore the best rack design for SBS/RS under
Our
CSP. aim
aim
We is
is to
to explore
explore
utilize the
the
simulation best
best
for rack
rack
the design
design
modelling for
for SBS/RS
SBS/RS
purpose. under
short response
Increasing times trend towardsa more
has created new AS/RS
more product
product variety
called SBS/RS
variety and
and under
short
short
(Carlo response
response
and Vis, times
times
2012; has
has created
created
Marchet a
a
et new
new
al., AS/RS
AS/RS
2012; called
called
Marchet SBS/RS
SBS/RS
et al., CSP.
CSP. We
CSP. We utilize
We utilize simulation
utilize simulation for
simulation for the
for the modelling
the modelling purpose.
modelling purpose.
purpose.
short response times has created a new AS/RS called SBS/RS
(Carlo
(Carlo and
and Vis,
2013; Lerher,
(Carlo and Vis,
Vis, 2012;
2012;
2013;
2012; LerherMarchet
et al.,et
Marchet
Marchet et al.,
al., 2012;
et2013).
al., 2012;
This new
2012; Marchet
Marchet et
et al.,
technology
Marchet et al.,
al.,
2013;
2013; Lerher,
Lerher, for
is developed
2013; Lerher, 2013;
2013;
2013;high Lerher
Lerher et
et al.,
al., 2013).
transaction
Lerher et al., 2013). This
This new
environments
2013). This new technology
newwhere
technology
mini-
technology
is
is
load
is developed
developed
AS/RS crane
developed for
for high
for high
highmay transaction
transaction
not be able
transaction environments
environments
to keep pace
environments where
where
where withmini-
mini-
the
mini-
load
load AS/RS crane may not be able to keep pace with the
load AS/RS
AS/RS
transaction crane
crane
rate may
may
needed not
not
overbe
be a able
able
given to
to keep
keep
number pace
paceof with
with the
storage
the
transaction
transaction
locations. rate
rate
Typically,needed
needed an over
over
SBS/RS a
a
transaction rate needed over a given number of storagegiven
given
is number
number
comprised of
of
of storage
storage
multiple
locations.
locations.
tiers Typically,
of storage
locations. with an
Typically,
Typically, an SBS/RS
SBS/RS
dedicated
an SBS/RSshuttlesis
is comprised
is for eachof
comprised
comprised of multiple
multiple
oflevel (see
multiple
tiers
tiers of storage with dedicated shuttles for each level (see
tiers
Figure of
of storage
storage
1). Hence, with
withloads dedicated
dedicated
are shuttles
shuttles
stored and for
for each
each
removed level
level
from (see
the
(see
Figure
Figure 1). Hence, loads are stored and removed from the
Figure
shelves 1).
1).
by Hence,
Hence,
SBS/RS loads
loadsat are
are
high stored
stored
speed, and
and
and removed
removed
the from
from
shuttles the
load
the
shelves
shelves by
handling
shelves by SBS/RS
byequipment
SBS/RS is
SBS/RS at
at high
high speed,
at designed
high speed,
speed, for andshortthe
and
and shuttles
thehandover
the load
shuttles times
shuttles load
load
handling
handling equipment
equipment is is designed
(http://www.dematic.com/multishuttle).
handling equipment is designed for
designed for short
for short handover
short handover times
handover times
times
(http://www.dematic.com/multishuttle).
(http://www.dematic.com/multishuttle).
(http://www.dematic.com/multishuttle).
A typical design of an SBS/RS is illustrated in Figure
A each
typical design of an
an SBS/RS
SBS/RS is illustrated
illustrated inShuttles
Figure Figure 1: An SBS/RS
1 whereA A typical
typical tierdesign
and aisle
design of
of anhas a shuttle
SBS/RS is and a lift.in
is illustrated in Figure
Figure Figure
1
1
arewhere
where
used each
each
as tier
tier and
and aisle
aisle
storage/retrieval has
has a
a shuttle
shuttle
devices and
and
in a
a
the lift.
lift. Shuttles
Shuttles
system to Figure 1:
Figure 1: An
1: An SBS/RS
An SBS/RS
SBS/RS
1 where each tier and aisle has a shuttle and a lift. Shuttles 2. LITERATURE REVIEW
are
are used as storage/retrieval devices in the system to
are used
used
store/retrieve as
as storage/retrieval
storage/retrieval
loads to/from devices
devices
storage in
in
locations. the
the system
system
Lifts provide to
to 2.
2. LITERATURE
2. LITERATURE REVIEW
LITERATURE REVIEW
REVIEW
store/retrieve
store/retrieve
vertical movementloads
loads to/from
to/from
for loads storage
storage
that are locations.
locations.
store/retrieve loads to/from storage locations. Lifts provide mostly Lifts
Lifts provide
provide
transferred by The earliest study on SBS/RS is completed by Carlo
vertical
vertical movement
movement
small containers (i.e.for
for loads
loads
fortotes). that
that are
that
loadsBecause are mostly
mostly
are each aisletransferred
transferred
has a single by
by
vertical movement mostly transferred by and VisThe
The earliest
earliest
(2012)
The study
study
where
earliest studytheyon SBS/RS
onstudy
on a typeis
SBS/RS
SBS/RS is ofcompleted
is completed
SBS/RS that
completed
by
by Carlo
by Carlo
there
Carlo
small
small
lift,
small containers
containers
it mostly
containers (i.e.
(i.e. totes).
(i.e. totes).
becomes totes). Because
Because
bottleneck
Because each
each
in theeach aisle
system.
aisle has
has a
aisleTherefore, single
has aa single
singlean and Vis (2012) where they study a type of SBS/RS that there
and
are Vis
two (2012) where
non-passing they
lifting study
systemsa type of
mounted
and Vis (2012) where they study a type of SBS/RS that SBS/RS
along that there
the rack.
there
lift,
lift, it
lift, it mostly
it mostly
alternative
mostly becomes
becomes
SBS/RS
becomes bottleneck
with ain
bottleneck
design
bottleneck inliftthe
in the system.
system. Therefore,
themechanism
system. Therefore,
having two
Therefore, an
an
an are two non-passing lifting systems mounted along the rack.
are
Theytwo non-passing
focus on lifting
scheduling systems
problem mounted
where along
two
are two non-passing lifting systems mounted along the rack. the rack.
(piece-wise
alternative
alternative
lifting tablesSBS/RS
SBS/RS
is design
design
developed. with
with
In aa lift
this lift mechanism
mechanism
design, the having
having
capacity of two
two
the They
alternative SBS/RS design with a lift mechanism having two
lifting They focus
They
linear)focus on
on scheduling
on
functions
focus scheduling problem
problem
are introduced
scheduling problem to where
where
where two
two (piece-wise
two
evaluate (piece-wise
candidate
(piece-wise
liftingaretables
lifting
lifts tables
tables is
is developed.
is
doubled developed.
by the two
developed. In
In this
this design,
this
In lifting tablesthe
design,
design, the capacity
capacityonof
capacity
attached
the of the
the
ofboth
the linear)
linear) functions are introduced to evaluate candidate
linear) functions
functions
solutions. are
are introduced
introduced to
to evaluate
evaluate candidate
candidate
lifts
lifts
side are
are
of doubled
doubled
the lifting by
by
system.the
the two
two lifting
lifting tables
tables
lifts are doubled by the two lifting tables attached on both attached
attached on
on both
both solutions.
solutions.
solutions.
side
side of
side of the
of the lifting
the lifting system.
lifting system.
system. Marchet et al. (2012) presented an analytical model to
estimateMarchet
Marchet et
et al. (2012)
et al.
al. (2012)
the performance
Marchet
presented
presented
(2012)(the
an
an analytical
an analytical
transaction
presented
model
model
cycle time
analytical
to
to
modeland
to
estimate the
estimate the
estimate performance
the performance
performance (the (the transaction
(the transaction cycle
transaction cycle time
cycle time and
time and
and
2405-8963 2015,
Copyright 2015 IFAC 1205Hosting by Elsevier Ltd. All rights reserved.
IFAC (International Federation of Automatic Control)
Peer review
Copyright
Copyright under
2015 responsibility
2015 IFAC
IFAC of International Federation of Automatic
1205Control.
1205
Copyright 2015 IFAC
10.1016/j.ifacol.2015.06.239 1205
INCOM 2015
B.Y. Ekren et al. / IFAC-PapersOnLine 48-3 (2015) 11521154 1153
May 11-13, 2015. Ottawa, Canada
waiting times) of SBS/RS. They developed an open queuing
network model whose performance effectiveness is validated
through simulation.
Later, Marchet et al. (2013) studied main design trade-
offs for SBS/RS using simulation. They complete their study
for several warehouse design scenarios for tier captive
shuttles. They provide several performance measures from
the pre-defined designs including cost.
Recently, Lerher (2013) and Lerher et al. (2013) have
studied energy efficient SBS/RS design. They considered
energy regeneration and energy efficiency concept in the
models. The proposed models enable reduction of energy
consumption and so CO2 oscillation, which is vital from
economic and environmental point of view.
None of the above studies consider CSP for SBS/RS
and compare the performance of the systems for different
rack design configurations. In this study, we consider CSP in
SBS/RS design and experiment several rack configurations to
find out the best one under this storage policy. The
performance of the system is evaluated in terms of utilization
of lifts and shuttles, average cycle time of storage and
retrieval transactions and, number of jobs waiting in queues
of lifts and shuttles as well as waiting times in these queues.
We utilize simulation for the modelling purpose and show
our results in a table.
3. SIMULATION MODELLING OF THE SBS/RS
The SBS/RS warehouse considered in the simulation
model consists of racks, lifts, and shuttles. Shuttles are tier-
captive and their main duty is to store/retrieve totes to/from
the storage locations. Racks (on either side of an aisle)
consist of bays, and each bay can hold one tote. Because each
aisle of the studied SBS/RS is identical in terms of number of
tiers and bays, we simulate a single aisle. The lifts are located
at the end of aisles and can carry two totes, independently.
Transaction seizes the available lift based on the dual
command (DC) scheduling rule. Each aisle has an I/O point.
The other assumptions that are used in our simulation model
are summarized below:
Arrivals follow a Poisson process and the mean
arrival rates for S/R transactions are equal and s =
R = 2,400 totes per hour.
The time for loading and unloading the totes to/from
shuttles/lifts are assumed to be 3 seconds each.
The shuttles and lifts acceleration and deceleration
delays are considered to be 2 m/sec2.
If the S/R transaction is located at the first tier then,
lift is not used.
The I/O location is located at the first tier of each
aisle.
There are two buffer positions at each tier where
lifting tables discharge and charge the loads.
The maximum velocities of shuttles and lifts are
considered to be 2 m/sec.
1206
We consider the velocity of lifts and shuttles as well
as the distance metrics of the warehouse from Lerher et al.,
(2013). The simulation model is completed using ARENA
14.0 a commercial simulation software and assumed to be a
non-terminating system, allowing us to conduct a steady state
analysis. The length of each simulation run is one year. The
warm-up period is defined to be three months by the eye-ball
technique. The model is run for 10 independent replications.
The verification of the simulation model is completed
by debugging the model via the trace module in the
simulation software. The validation of the model is
completed by comparing our average travel time of lifts with
the analytical model results in Lerher et al., (2013).
3.1 Scenarios for Simulation Runs and Results
Simulation runs are conducted based on the Table 1
scenarios. In this table, T, B, A, Q stand for number of tiers,
bays, aisles and number of storage locations in the
warehouse, respectively. For the Q value, we assume that
based on the maximum space utilization of racks at a time
10,000 number of storage locations are required. Based on
Table 1, we consider seven T values: 14-20, and four A
values: 5-8. We calculate the B values so that the
multiplication of T, B, and A provides half of the Q value
(due to the two sided storage columns). For instance, in the
first scenario of Table 1, T and A values are defined to be 14
and 7, respectively. Hence, the B value is obtained to be 51
by rounding to the nearest integer (5,000/(147)). In SBS/RS,
because each tier has a shuttle, lifts usually become
bottleneck. The scenarios of T, B and A are defined by
considering reasonable utilization values (i.e., mostly
between 60% - 95%) of the bottleneck resources (i.e.,
shuttles) to be obtained.
The scenarios based on T, B and A are implemented
for the CSP in the SBS/RS. Storage-location assignment in
warehouses is an important task since it affects productivity
of other warehouse processes. The CSP policy distributes the
products, among a number of classes, and for each class it
reserves a region within the storage area. In CSP A-class of
items are usually stored at the closest available location.
Hence, the cycle times are usually low and so the throughput
rates are high compared to a random storage policy. In the
simulation model, we implement the CSP as follows:
In the class-based storage policy, the items are
categorized into three classes A, B, and C. Because lifts are
mostly bottlenecks in the system, each class is assigned to a
dedicated tier(s) in the warehouse. For instance, class A items
are stored at the closest tiers to the I/O-point and class C the
farthest. Particularly, in the simulation model, we create A
class of items with 80% probability and assign them to the
first 20% of the total tier. For instance, if there are 14 tiers in
a scenario, with 80% probability (A-class items), an item is
stored at the first 3 tiers (20% 14) randomly; with 15%
probability (B-class items) it is stored at the following 30%
tiers randomly (30% 14) - tiers 4-7 -, and with 5%
probability it is stored at the remaining tiers: 8-14, randomly.
INCOM 2015
1154 B.Y. Ekren et al. / IFAC-PapersOnLine 48-3 (2015) 11521154
May 11-13, 2015. Ottawa, Canada

We assume that the storage bays within the designated tiers Based on the utilization values, the plant manager
are assigned to be randomly. may prefer the design with 20, 42, 6 for T, B, A,
respectively. In this scenario, the utilization values
Recall that the performance of the system is measured of lifts/shuttles are around 75%.
in terms of utilization of lifts (U_L) and shuttles (U_S),
average cycle time of storage (C_S), and retrieval (C_R) 4. CONCLUSIONS
transactions. The simulation results are summarized in Table
1. It should be noted that the results provided in this table are In this study, a simulation based warehouse design
the average values of ten replications and the C_S and C_R analysis for SBS/RS is presented. We believe that by this
values are measured in minute unit. The simulation results study, we provided significant design insights for SBS/RS for
are obtained at 95% confidence level. In Table 1, the both practitioners and academics. In the study, the warehouse
performance measure U_S is provided for shuttles serving design is considered in terms of number of tiers, bays and
class A. This is because these shuttles become bottleneck in aisles in the rack configuration. We explore the best rack
the system. design for class-based storage policy. We evaluate the
performance of the system for several rack designs. The
Table 1: SBS/RS scenarios and simulation results performance of the system is evaluated in terms of
T B A Q U_L U_S C_R C_S utilizations of lifts and storage/retrieval devices and cycle
14 51 7 9996 Blocked Blocked Blocked Blocked times of storage/retrieval transactions. We utilize simulation
14 45 8 10080 0.56 0.96 3.46 3.49 for the modelling purpose. Simulation results are summarized
14 40 9 10080 0.43 0.65 0.59 0.63 in a table. Based on this table, it is observed that CSP works
15 56 6 10080 Blocked Blocked Blocked Blocked well under high rise warehouse design than low rise
15 48 7 10080 0.56 0.93 2.05 2.08 warehouse design. Within a fixed tier scenario, the
15 42 8 10080 0.49 0.75 0.75 0.79 performance of the system becomes better when the number
16 52 6 9984 Blocked Blocked Blocked Blocked of aisles increases (in this case bays decreases).
16 45 7 10080 0.56 0.89 1.46 1.50
16 39 8 9984 0.49 0.73 0.68 0.72 This study can be extended in many directions. For
17 49 6 9996 Blocked Blocked Blocked Blocked instance, the design scenarios may be extended by including
17 42 7 9996 0.57 0.86 1.20 1.18 different arrival rate scenarios as well as velocity profiles for
17 37 8 10064 0.50 0.71 0.64 0.67 lifts and shuttles. Besides, different storage policies (e.g.,
18 56 5 10080 Blocked Blocked Blocked Blocked random storage policy, closest available location, etc.) as well
18 46 6 9936 0.74 0.79 0.94 0.97 as operational policies (scheduling rules, shuttles that can
18 40 7 10080 0.63 0.63 0.60 0.64 travel between tiers, etc.) may also be considered in the
19 53 5 10070 Blocked Blocked Blocked Blocked
design scenarios and compared with each other.
19 44 6 10032 0.74 0.77 0.87 0.90
19 38 7 10108 0.63 0.61 0.58 0.62
REFERENCES
20 50 5 10000 0.89 0.99 16.59 16.37
20 42 6 10080 0.74 0.75 0.81 0.84 Carlo H.J. and I.F.A.Vis (2012). Sequencing dynamic storage
20 36 7 10080 0.63 0.60 0.56 0.59 systems with multiple lifts and shuttles. International
We summarize our observations from Table 1 as in Journal of Production Economics, 140, 844-853.
below: Lerher, Tone, Edl, Milan, Rosi, and Bojan. (2013). Energy
efficiency model for the mini-load automated storage
From Table 1, we observe that shuttles serving for and retrieval systems. International Journal of Advanced
A-class items tend to become bottlenecks in the Manufacturing Technology, 70(1-4), 97-115.
system. Lerher, Tone. (2013). Modern automation in warehousing by
When we consider T as 18-20, all the performance using the shuttle based technology. V: ARENT, Doug
measure values are typically better than the ones in (editor), FREEBUSH, Monica (editor). Automation
scenarios having 14-17 tiers. Hence, we conclude Systems of the 21st Century: New Technologies,
that CSP works better under high rise warehouse Applications and Impacts on the Environment &
design than low rise design. Industrial Processes. New York: Nova Publishers, cop.
In Table 1, the system is blocked when the number 51-86.
of aisles is low in the warehouse. This is probably Marchet, G., Melacini, M., Perotti, S. and Tappia, E. (2012).
because, under this condition the arrival rate to an Analytical model to estimate performances of
aisle increases (which is calculated by 2,400 autonomous vehicle storage and retrieval systems for
transactions/h / A). Under this condition to obtain product totes. International Journal of Production
the required Q value, the number of bays tends to Research, 51(14), 43654387.
increase. As a result of this, shuttles serving A-class Marchet, G., Melacini, M., Perotti, S., Tappia, E. (2013).
of items become bottleneck. It should be noted that Development of a framework for the design of
in a scenario within a constant T, the performance of autonomous vehicle storage and retrieval systems.
the system becomes better when A increases (in this International Journal of Production Research, 50(24)
case B decreases). 71347148.

1207