OPERATIONS BY FORKLIFTS IN WAREHOUSES

Aurelija Burinskiene

Vilnius Gediminas technical university

aurelija.burinskiene@vgtu.lt

ABSTRACT warehouses. The movement of a pallet from point A to

Nowadays the great attention has to be paid to logistics. point B is not critical itself, but if the wrong pallet is
Theoretical analysis shows the need to improve logistics moved or if the right pallet is put in the wrong place, an
system. This encourages the search for solutions during error will occur.
warehouse operations. Aiming to determine possible During the process, driver have to know where to
improvements, the operations of forklifts are examined go, how much to pick and where to place it. The
in warehouses. construction of least-costs forklift routes can help to
The study consists of two parts: in the first part of the increase productivity.
study possible improvements are analysed, in the Forklifts are used in environments and applications
second part - the financial evaluation of suggested that require different performance characteristics than
improvements are presented. other equipments.
The application of simulation method is presented in the Forklifts retrieve products from specified storage
first part of the study, the use of the comparative and locations on the basis of customer orders. The
financial analysis methods - in the second part of the infrastructure needed for a firm to incorporate and
study. handle all movements of forklifts has to be in place.
Simulation results were compared with the results of Forklifts are the most expensive machines due to
companies, applying such solutions. This showed labor investments into equipment and other assets, which are
productivity increased from 27% to 37%. Continuing used to support their operations.
the investigation, the assessment of financial value is This shows the importance of such research. In
presented. The solution describing how to improve addition, the analysis of scientific literature, which is
logistics system, gives time savings around 12.5% and published by Oxford University Press, Cambridge
cost savings around 9%. University Press, Harvard University Press, Springer,
The presented results resemble to the results reached in M.E. Sharpe, Routledge, etc. shows that only 1% of
practise. authors, which analyze logistics issues, give attention to
forklifts. The authors that analyze the operations of
Keywords: routing, warehouse, simulation, forklifts, forklifts mention that minimization of non-productive
savings movements and the construction of less-costly routes
can help to increase productivity.
1. INTRODUCTION In the paper author analyzes the application of
Warehousing is becoming critical activity in trade forklifts in trading companies, business cases, the
enterprises to outperform competitors on productivity, assessment of technology implementation by production
lead time and customers’ service. So, short lead times companies, their attitudes and interests.
and productivity are considered in many warehouses as The target of the paper is to determine possible
competitive weapon. It becomes more and more logistics systems improvements in forklifts operations.
difficult to realize short customers’ orders throughput To answer to the above mentioned question, study
times because of factors such as increase in assortment is presented in the paper.
and smaller, not frequent, orders. For the increasing The study is divided into two parts. At the
assortment an increasing floor space is needed. This beginning literature review is given, later on
means in increased lead times per order. Smaller orders experimental study is presented. The literature review is
(less units per line) and higher frequency of ordering short summary of the application of forklifts in trading
lead to some changes: less full pallets are picked and companies, business cases, the assessment of
more single item picks are necessary. Also additional technology implementation by production companies,
activities have to be performed within the short time their attitudes and interests. The first part of
frame available for handling the customers’ orders. experimental study is dedicated for the identification of
For such activities forklifts are the most important forklift and construction of least-costs forklift routes.
and expensive machines. Requirements for forklifts and The second part of experimental study is dedicated for
costs are crucial for the selection of such operations. the effect related to minimization of costs identification
Basically, forklifts are used for moving pallets inside and time savings calculation.
 The study is based on simulation, comparative and to storage area. At the end this means increased
financial analysis methods. throughput time per order.
According to De Koster, Roodbergen, and Voorden
2. THE SCOPE OF FORKLIFTS APPLICATION (1999), the travel time in a warehouse, in general, has a
IN WAREHOUSES significant impact on throughput time.
When purchasing forklift, the purchaser must know, if it The time, needed for forklifts operations can be
can meet handling needs of company. This is where split in travel time (travelling time), pick and place
productivity comes in. Speed limits are applicable and time, and remaining time. The travel time is required for
they have huge impacts. Usually manufacturers of the movement between locations that have to be visited
forklifts can design governors on forklifts that only (where the items are stored that have to be moved). The
allow reaching the certain speed (Ryan and Ryan 2006). pick time is associated with stop, finding and picking
The speed limit in different warehouses is 10km/h. right pallet. This includes search for the pallet, grabbing
A forklift is a powerful tool when used by well it, scanning it and storage location, reading the next
trained driver (or operator). It helps to move pallets (or location to be visited. Place time is related to the
other materials) by eliminating the need to lift and carry dropping of the pallet to pick location and scanning it.
boxes by hands. The remaining time is related to time needed to close
Forklifts can be classified by different methods the task and get next assignment, etc.
including type of power source, driver position, and
load engagement method. Electric motors powered by 2.1. Time savings algorithms
batteries are among the most common type of forklifts 50% of the total travel time spent on unproductive
seen in industry (Ryan and Ryan 2006). Different traveling (Dukic and Oluic 2007). Saving algorithms are
manufacturers may offer different models of forklifts. based on the algorithm of Clarke and Wright (1964) for
Selection of forklift is often a several-step decision the vehicle routing problem: a saving on travel distance
process. First characteristic is volume of material flow - is obtained by combining a set of small tours into a
the large volume will require buying the number of smaller set of larger tours.
forklifts with similar characteristics. Second Time savings algorithms are based on the time
characteristic is the frequency and scheduling of savings that can be obtained if different tasks are
movements - this relates to the time between combined instead of being executed separately. For
movements. Usually, movements are performed due to combined tasks the time saving is calculated.
different batches or continuous. Third characteristic is Instructions for the next route are given via a computer
route factor - this includes distance, conditions along terminal (De Koster, Le-Duc, and Roodbergen 2006).
the route, the path of travel (Kasilinga 1998). For reaching next assignment, radio data terminal is
In addition, author Kasilinga (1998) mentions used by forklift driver. The forklift driver follows a
important forklift selection factors: purchasing price, route designated by reading instructions on a radio data
operating costs, maintenance costs. The regulation terminal or following commands.
require that a forklift be checked every day that the Each route is formed containing many tasks. This
forklift is used. If forklift is used continuously, then it process is repeated until all tasks during day have been
must be checked at the end of each shift. assigned to a route.
Forklift operations require the floor space as well. In the calculation of the time savings, a routing
Traditional wide-aisle pallet storage aisles must have algorithm may be used. This may be one of mentioned
corridors 3.3-3.6 m for forklifts operations. It is heuristics: S-shape, Combined, Largest gap, return,
necessary to have also barriers that protect walls and mid-point, etc. or optimal (De Koster, Le-Duc, and
racks against forklift damages. Roodbergen 2006). Usually optimal method
The place of depot in front of aisles or between the outperforms S-shape method by 7-33 % in single-block
two blocks is planned in Caron, Marchet, and Perego random storage warehouses (De Koster and Van der
(1998) study. This is the position where forklift driver Poort 1998). Several types of the above mentioned
takes task and after which returns back for another one. algorithms are discussed in De Koster, Roodbergen,
Various assignment policies may be used and Voorden (1999).
simultaneously within the area of warehouse. Consider, All articles discussed so far assume that the aisles
for example, picking and storage areas. The used of the warehouse are narrow enough allowing to
assignment policy does not have to be permanent and retrieve pallets from both sides of the aisle without
could be switched to another one when conditions are changing position. In Goetschalckx and Ratliff (1988) a
changed. Dedicated storage policy can be used for fast polynomial-time optimal algorithm is developed that
moving products and random storage policy for slow solves the problem of routing order pickers in wide
moving items (Hassan 2002). aisles.
The number of locations available in storage area In practice, usually simple routing heuristics are
usually is smaller than the total number of pallets used. No standard warehouse management system
arriving to warehouse on daily basis. This means that package consists that contains the Ratliff and Rosenthal
first, pallets from storage places have to be removed to algorithm. Methods that are used include the ‘Largest-
picking places and after this, new pallets can be placed gap’ return strategy, ‘Mid-point’ return strategy, ‘S-
shape’ strategy. In the application of ‘S-shape’ costs forklift routes. The second part of experimental
algorithm, a distinction has to be made between single- study is dedicated for the effect related to minimization
side and double-side picking (De Koster, Roodbergen, of costs identification and time savings calculation.
and Voorden 1999). For the first part of experimental study the
In many labor-intensive warehouses, there is the Interactive warehouse simulation model is used.
constant pressure to reduce travelling time. One of such The Interactive Erasmus Logistica Warehouse
is Bijenkorf warehouse - warehouse of retail trade Website model offers opportunities to discover more
organization in Holland with the product assortment of ways to perform tasks (Oudijk, Roodbergen, De
300 thousand stock-keeping units. Authors De Koster, Koster, and Mekern 2002) in the Microsoft Internet
Roodbergen, and Voorden (1999) have made the case Explorer environment. There are different possibilities
study analysis. The delivered results are as follows: the to compare different scenarios which depend from the
reduction of total travel time means the reduction in the number of aisles and cross-aisles, length of aisles,
number of labor units. position of computer station in warehouse. The only
Another problem with routing may arise if products limitation of the model is that the maximum number of
are stored at multiple locations in a warehouse (De locations per aisle is limited to 68.
Koster, Le-Duc, and Roodbergen). In this case a choice The Interactive warehouse simulation model will
by warehouse management system has to be made be used for optimizing the way of forklift that is done
before directing the forklift to location from which the between two transaction points. Each time forklift gets a
pallet has to be retrieved. task, an event (or task) row is recorded in the
The travel time is an increasing function of the information system. An event row can include the
travel distance (study, for example, Jarvis and starting and ending time of task, thus enabling its
McDowell 1991; Roodbergen and De Koster 2001; duration to be calculated. Each consecutive pair of
Petersen and Aase 2004). Consequently, the travel transaction points represents costs of activity (Varila,
distance is often considered for planning warehouse. Seppanen, and Suomala 2007).
Two types of travel distance are widely used in The forklift driver starts route at computer station
appropriate literature: the average travel distance of tour (Oudijk, Roodbergen, De Koster, and Mekern 2002),
(or average tour length) and the total travel distance. For goes to the front of the particular aisle; drives inside
a given tasks list (a set of tasks), however, minimizing aisle, takes pallet, delivers it to pick location at the same
the average tour length is equivalent to minimizing the or other aisle in shortest way, and, finally, returns back
total travel distance. to the computer station. For calculation of forklift travel
Clearly, the minimization of average travel distance distance the optimal route was used mainly (see Figure
(or, equivalently, total travel distance) is limited for 1). Each pick and place point is served by only one
forklift operations. Usually, it is only one of many forklift. In addition, the capacity of forklift is limited. It
possibilities. Another important objective would be is possible to handle one pallet on its forks at the same
minimizing the total cost (De Koster, Le-Duc, and time.
Roodbergen 2006) (that may include both investment in Three assumptions are used for experimental study.
to forklifts and related infrastructure and operational The first one is dedicated to optimal routing by
costs). assuming that order pickers do not create routes on their
In addition tracking system maybe used that will own. The next one is dedicated to the wide of the aisle.
connected forklifts to the mobile device for the This means that the aisle has the least floor space where
provision of routing information (Giaglis, Minis, the forklift can pick pallets from both sides of an aisle
Tatarakis, and Zeimpekis 2004). without covering an additional distance. The third one is
Eastern Service Company - freight forwarding and dedicated to path and time relationship by assuming that
logistics service provider in Hong Kong and South the linear relationship between forklift travel time and
China. The company specializes in managing distance exists. For the calculation of traveling time the
warehouses and providing other cargo logistics services. average speed 10 km/h is used (assuming that forklift is
This company aims to achieve these major service goals travelling in warehouse with the same speed).
to their customers: shorten the lead time; and maintain a This simulation model was used to perform pick
higher visibility on the supply chain area. To ensure and place tests, when:
accuracy scanning is applied to check the right 1. Reserve (pick) and pick (place) locations were
correctness of movement (Cathy, Lam, Choy, and in different aisles (see the left side of the
Chung 2011). picture);
2. Reserve and pick locations were in the front of
3. EXPERIMENTAL STUDY aisle;
The objective of experimental study is to calculate how 3. Reserve and pick locations were in the end of
the identification of forklift will help for the aisle;
construction of least-costs forklift routes; and what will 4. The above mentioned is combined (see the
be the effect related to minimization of costs. middle of the picture).
The first part of experimental study is dedicated for
the identification of forklift and construction of least-
 Figure 1: Interactive warehouse simulation model and
different tests. Source: Oudijk, Roodbergen, De Koster,
and Mekern (2002)
Figure 2: One return to depot for two tasks. Source:
Oudijk, Roodbergen, De Koster, and Mekern (2002)
If forklift works according paper-based task list, at
first forklift driver moves to depot (where all task lists
are placed), then moves to storage location to pick pallet In addition, authors Gray, Karmarkar, and
(according task list), takes pallet, delivers it to pick Seidmann (1992) propose to analyse savings in
location, and, finally, returns back to the depot to accordance with the components of warehouse
confirm completeness of task and gets assignment for operations. The time, needed for forklifts operations,
new task. The trip to depot takes some minutes. can be split in travel time (travelling time), pick and
If forklift has RF terminal, then new task list is place time, and remaining time:
placed on the screen (no need to drive to depot). In such  The travel time is required for the movement
case the driver of forklift has to be equipped with radio between locations that have to be visited,
control device, such as RF terminal, which is located on  The pick time is associated with the number of
the forklift. The terminal of forklift can have real-time stops, finding and picking right pallet,
communication with warehouse management system.  Place time is related to the dropping of the
And the portable computer can be used to map out the pallet to the right location,
most efficient route. Operations performed by forklifts  The remaining time is related to time needed to
with RF terminal require a discrete signal that will close the task and get next assignment at
ensure only that the desired pallet is identified. RF computer station.
terminal is used to send messages (SME’s) for the After each above mentioned component is studied:
driver. The monitor on the forklift typically provides  Travel time component. Gray, Karmarkar, and
instructions associated with the given task. When the Seidmann (1992) propose to calculate 1,5
task is finished the driver gives confirmation to minutes per performed task. On average during
warehouse management system over local area wireless a day he spends an hour and a half for
(WI-FI) network (see Figure 2, when several or multiple travelling in order to take a pallet and to place
tasks are completed without return to depot). a pallet into new location. Comparing results
By running simulation tests with interactive of two scenarious (Figure 1 and Figure 2) it is
warehouse simulation model it was seen that the best seen that the total travel distance of forklift can
travel distance is when RF-terminal is implemented. be reduced by 27-37 %;
This means that the tasks are distributed to forklifts via  Pick time component. Gray, Karmarkar, and
SME’s and forklift drivers don’t need to go to depot to Seidmann (1992) propose to calculate 1 minute
give confirmation and get new assignments. In such per pallet taken from the old place (0.5 minute
case, the way decreases by 27-37 % when travelling of this is needed for finding the right pallet and
without depot is implemented and tasks are distributed scanning location). If on average the driver of
to forklift drivers on-line. forklift picks 60 tasks, this means that during a
Talking about costs, these results present the cases day he spends around an hour to pick pallets;
with and without RF-terminals. As there are several  Place time component. Gray, Karmarkar, and
choices for warehouses - task lists distributed to forklift Seidmann (1992) propose to calculate 1.5
drivers at depot (computer station) or via SME’s. minute per pallet placed to new place. On
average during a day the driver of forklift
spends one hour and a half to place pallets;
 Remaining time component. In paper-based
process it takes some forklift driver’s efforts to
 handle the job at the depot (computer station).
Gray, Karmarkar, and Seidmann (1992) The mentioned numbers are presented in Table 2.
propose to calculate 0.5 minutes per task
(Gray, Karmarkar, and Seidmann 1992). If on Table 2: Total investments required for forklifts
average the driver of forklift picks 60 tasks, operations
this means that during a day he spends around
Investments 1st 2nd
half an hour at the computer station.
scenario scenario
The mentioned time components for the first and
second scenario are mentioned in Table 1.
Investments into 26106 21106 26106
forklift Euro Euro Euro
Table 1: The time needed for forklift operations
Components 1st 2nd Effec Warehouse space (50 1000
scenario scenario t m x 20 m) sq. m

Travel time 1.5 1.5 Charging room space 7.4 sq.

component meter

Pick time 1 1 Investments into 4526 4526 4526

component charging room Euro Euro Euro

Place time 1.5 1.5 Rent price per sq. 5

component meter Euro

Remaining time 0.5 0 Number of corridors 10.5

component with 3.3 m width and
20 m length
Total time 4 3.5 12.5
% Investments into 6740 3340 6740
warehouse floor Euro Euro Euro
The effect 12.5% related to minimization of costs is Forklift driver costs a 40 40 Euro 35 Euro
identified (see Table 1). day Euro
And finally, time savings calculation is presented.
First, each line of investments is mentioned, later on the Total costs for 5 78372 80597
comparison for both scenarios is given. years Euro Euro
Investments into forklift: Savings in 5 years
The forklift purchase price - 18113 Euro
The purchase price of 2 batteries - 2993 Euro Savings of driver 6175
The purchase price for RF-terminal* - 5000 Euro costs (when 247 Euro
Total purchase price for forklift and its equipments - working days a year)
26106 Euro
Savings of forklift 3263
Investments into charging room: costs (12.5%) Euro
The purchase price for charging room protectors - 23
Euro per sq. meter Total savings 9438
The purchase price for battery change table - 1496 Euro Euro
The purchase price for 2 battery chargers - 2860 Euro
Net total costs for 5 78372 71159
Investments into warehouse floor: years Euro Euro
The purchase price for walls and racks protectors - 2.5
Euro per sq. meter
In addition, it should be mentioned that net total
The purchase price for WI-FI network* - 3.4 Euro per
sq. meter costs for second scenario are 91 % lower than for the
first scenario. This means that if more forklifts will be
Increase of corridor space from 2.5 sq. meter to 3.3 sq.
meter for forklifts operations included in such calculation, the reduction of travel
distance effect would be much higher.
One sq. meter of charging room is requested for 135 sq.
meter of warehouse floor
4. CONCLUSIONS
* - additional investments needed for second scenario. In many labor-intensive warehouses, there is the
constant pressure to reduce travelling time. In the
literature review short summary of the application of literature review. ERIM Report series research in
forklifts in trading companies, business cases, the management.
assessment of technology implementation by production Dukic, G., Oluic, C., 2007. Order-picking methods:
companies, their attitudes and interests. In addition, improving order-picking efficiency. International
different time savings algorithms are presented. journal of logistics systems and management, 4(3),
In the second part of the study is dedicated to 451-460.
experiments. First, to construction of least-costs forklift Giaglis, G.M., Minis, I., Tatarakis, A., Zeimpekis, V.,
route; and second, to effect, related to minimization of 2004. Minimizing Logistics Risk through Real-
costs, estimation. Time Vehicle Routing and Mobile Technologies.
The experiments demonstrate labor productivity International Journal of Physical Distribution &
increase from 27% to 37% and reduction of time and Logistics Management, 34, 749-764.
costs. The comparison results of both scenarios shows Goetschalckx, M., Ratliff, D.H., 1988. Order picking in
that total time can be reduced by 12.5 % due an aisle. IIE Transactions, 20, 531-562.
construction of least-cost forklift routes. In addition, Gray, A.E., Karmarkar, U. S., Seidmann, A., 1992.
total costs can be reduced by around 9%. Design and operation of an order-consolidation
In the paper the travelling of forklift is examined warehouse: models and application. European
through simulation and experiments with simulation Journal of Operational Research, 58, 14-36.
model are undertaken. For saving analyze, the time, Hassan, M.M.D., 2002. A framework for the design of
warehouse layout. Facilities, 20(13/14), 432-440.
needed for forklifts operations, is splited into travel
Jarvis, J.M., McDowell, E.D. 1991. Optimal product
time, pick and place time, and remaining time. The layout in an order picking warehouse. IIE
results of such analysis showed that the yearly savings Transactions, 23(1), 93-102.
for each forklift driver are 1235 Euros and for each Kasilinga, R.G. 1998. Logistics and transportation -
forklift - 652 Euros. design and planning. USA: Springer.
In addition, it should be mentioned that net total Petersen, C.G., Aase, G., 2004. A comparison of
costs for second scenario are 91 % lower than for the picking, storage, and routing policies in manual
first scenario. This means that if more forklifts will be order picking. International Journal of Production
included in such calculation, the reduction of travel Economics, 92, 11-19.
distance effect would be much higher. The study results Oudijk, D., Roodbergen, K.J., De Koster, R., Mekern,
show that the implementation of RF-terminals and M., 2002. Shelf Area Warehouse Simulation.
related infrastructure is a priority in warehouses. By Erasmus University Rotterdam and Stichting
implementing this managers can have significant Logistica. Available from:
savings of warehouse costs, especially for forklifts, http://www.roodbergen.com/warehouse [accessed
which are the most expensive machines used for manual 24 November 2010]
operations in warehouses. Roodbergen, K.J., De Koster, R., 2001. Routing
methods for warehouses with multiple cross aisles.
REFERENCES International Journal of Production Research,
Caron, F., Marchet, G., Perego, A., 1998. Routing 39(9), 1865-1883.
policies and COI-based storage policies in picker- Roodbergen, K.J., De Koster, R., 2001. Routing order-
to-part systems. International Journal of pickers in a warehouse with a middle aisle.
Production Research, 36(3), 713-732. European Journal of Operational Research, 133,
Cathy, H.Y., Lam, K.L., Choy, S., Chung, H., 2011. A 32-43.
decision support system to facilitate warehouse Ryan, J.L., Ryan, L.D., 2006. Forklift manual. USA:
order fulfillment in Donegal Bay Inc. publishing.
cross-border supply chain. Journal of Manufacturing Varila, M. Seppanen, M., Suomala, P., 2007. Detailed
Technology Management, 22(8), 972 - 983. cost modelling: a case study in warehouse
Clarke, G., Wright, W., 1964. Scheduling of vehicles logistics. International Journal of Physical
from a central depot to a number of delivery Distribution & Logistics Management, 37(3), 184-
points. Operations Research, 12, 568-581. 200.
De Koster, R., Van der Poort, E.S., 1998. Routing
orderpickers in a warehouse: a comparison AUTHORS BIOGRAPHY
between optimal and heuristic solutions. IIE Aurelija BURINSKIENĖ Dr.Soc.sc., Employee of
Transactions, 30, 469-480. Vilnius Gediminas technical university, Lithuania; e-
De Koster, R., Roodbergen, K.J. and Van Voorden, R., mail: Aurelija.Burinskiene@vgtu.lt. Her research is
1999. Reduction of walking time in the focused on ICT and e-commerce. Her job background
distribution center of De Bijenkorf. In: M. G., covers thirteen years’ job experience at wholesale and
Speranza, P. Stähly, eds. New Trends in retail trade logistics and supply chain; for last six years
Distribution Logistics, Berlin: Springer, 215-234. the most of supply chain projects have covered Baltic
De Koster, R., Le-Duc, T., Roodbergen, K.J., (2006). countries.
Design and control of warehouse order picking: a