An Analysis and Design for the Repair Process of Late Show
Shipments in the Export Cargo Process at SPL HUB
Sjoerd van Rooden
1
, Catya Zuniga
1
, Bart Krol
2
and Elias Olivares-Benitez
3a
1
Faculty of Technology, Amsterdam University of Applied Sciences, Weesperzijde 190, Amsterdam, The Netherlands
2
KLM Cargo, Schiphol, The Netherlands
3
Facultad de Ingenieria, Universidad Panamericana, Zapopan, Mexico
Keywords: Late Shipments, Late Shows, Repair Process, Acceptance Process, Business Process Modelling, Air Cargo
Operations.
Abstract: Export shipments arriving late at the freight building of KLM Cargo at Schiphol Airport is a trigger to
deviations in the standard acceptance process. These Late Shows are currently handled ad-hoc making it
difficult to plan and predict these events. By conducting a data analysis to quantitatively identify the
characteristics of the Late Shows, and by conducting stakeholder interviews to understand the current process
and discuss the future process, this research tried to design the operational process of the Late Shows to
improve the operational excellence and quality of the acceptance process. The research shows that currently,
late shipments are often still tried to be build up for the planned flight. It is found that 13% of these shipments
do eventually not depart on the planned flight. The research concludes that the design of the Late Show process
should include a check on whether the shipment was delivered on time, before acceptance of the shipment.
By only accepting the shipment once it is decided that the planned flight is achievable or when it is rebooked
to another flight, it is assured that the Late Show will be on time at the build-up buffer for the booked flight.
1 INTRODUCTION
The analysis of airport operations is of great
importance to improve the efficiency and quality of
the processes (Blonk, 2017; Henriksson & Petersson,
2019). KLM Cargo found that export shipments
arriving late at their freight building at Schiphol
Airport (SPL) is a trigger to deviations in the standard
process of accepting shipments from forwarders. Late
shipments are defined as shipments that are unloaded
from the truck after the latest acceptance time and
before flight departure. These shipments are called
“Late Shows”. In order to create more time for the
ground processes resulting in a reduced chance of
shipments missing their flight, KLM Cargo
implemented two changes to the acceptance process
on January 11, 2021. First, the Freight on Hand
(FOH) moment is moved from the Documentation
station to the warehouse. Thus, the moment of
acceptance of the shipment is no longer when the
driver reports at the Documentation station, but when
the shipment is unloaded and available in the
a
https://orcid.org/0000-0001-7943-3869
warehouse. Second, the latest time before flight
departure that shipments delivered by forwarders are
accepted, the Latest Acceptance Time (LAT), is
increased. The new LAT’s differ per product type.
Currently, there is no standard process for
shipments arriving late, meaning they are being
handled ad-hoc. Therefore, it is not possible to plan
these events, causing the process to be less
predictable, which negatively influences the
operational excellence of the process. In addition,
shipments arriving on time is currently not a criterion
for acceptance. However, when the shipments are
accepted, they should be able to catch the flight on
which they are booked, or the quality of the process
deteriorates. This can be explained by the definition
of “quality” and the definitions of two messages in
the standard acceptance process. This research
considers the definition of Slack, Brandon-Jones, &
Johnston (2016), who define quality as “consistent
conformance to customer expectations” (p. 573).
Further, IATA defines two messages, or status events,
in the acceptance process. IATA mentions that after
van Rooden, S., Zuniga, C., Krol, B. and Olivares-Benitez, E.
An Analysis and Design for the Repair Process of Late Show Shipments in the Export Cargo Process at SPL HUB.
DOI: 10.5220/0010797500003117
In Proceedings of the 11th International Conference on Operations Research and Enterprise Systems (ICORES 2022), pages 123-130
ISBN: 978-989-758-548-7; ISSN: 2184-4372
Copyright
c
2022 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
123
the truck is unloaded, the “Freight on Hand” message
is sent. This message indicates, “The consignment is
on hand on this date at this location pending ‘ready
for carriage’ determination.” (IATA, 2019, p. 1).
Then, after the necessary checks for safe and secure
carriage are conducted and the shipment information
is validated against the booking, the “Received from
Shipper” (RCS) message is sent. This message
indicates, “The consignment has been physically
received from the shipper or the shipper’s agent and
is considered by the carrier as ready for carriage on
this date and this location.” (IATA, 2019, p. 1).
Considering these definitions of IATA, it is important
to understand the difference between the acceptance
of shipments in terms of receipt in the warehouse
(FOH) and the acceptance of shipments as ready for
carriage (RCS). Further, it can be understood from
these definitions that a shipment should depart on the
booked flight at the moment of acceptance (RCS).
Whether the shipment departed on the flight on which
it was booked at acceptance is indicated by the DEP-
R.
The objective of this research is to design the
collaborative operational process for repair of the
Late Shows in the export acceptance process of KLM
Cargo at SPL. This should result in a DEP-R
improvement of 0.5% and 98% of the shipments
should be delivered on time at the buffer for buildup.
The main research question is formulated as: How
can the process of export CARGO acceptance be
designed in such a way that shipments will be on time
at the buffer for buildup for the booked flight?
The research is focused on the acceptance of
shipments delivered to freight building 3. Further, the
research is conducted before and during the
implementation phase of the Late Show process for
the first product types.
2 METHODOLOGY
The research is approached using a combination of
both quantitative and qualitative methods. First, a
data analysis (desk research) is conducted using
quantitative methods to identify the characteristics of
the Late Shows. This data is collected using two data
sources. One of these sources is the leading source for
the analysis and the other is used as backup and to
validate the data collected from the first source. The
data is exported into Excel-sheets, after which it is
analyzed using pivot tables, graphs and other
(statistical) functions. The data analysis exists of a
statistical analysis into the predictability of the
number of Late Shows and total shipments and more
detailed analyses into the characteristics of the Late
Shows. The statistical analysis is conducted on four
levels: the number of shipments per month, per week,
per day of the week and per day. On each of these
levels, the number of shipments during the analyzed
Figure 1: General acceptance process at KLM Cargo SPL.
ICORES 2022 - 11th International Conference on Operations Research and Enterprise Systems
124
period of January, February and March 2021 is
plotted in boxplots. This figure presents the
distribution of the data points (number of shipments),
giving information such as the mean, median,
minimum, and maximum value in the data set. For the
number of Late Shows, another level is analyzed,
presenting an overview of the Late Shows during
each hour of the day.
In the second phase of the research, it was tried to
understand first the current (ad-hoc) handling of the
late shipments. The data used to map this current
process is collected using interviews with the
stakeholders of the process (field research). During
these interviews, the new Late Show process, that
already had been tested, was also discussed with the
stakeholders. They were asked about their opinion on
the new Late Show process as well as already
considered possible improvements for the process, as
identified during process test evaluations. A
representative (or sometimes two) of each involved
department is selected for the interview, chosen
because of their expertise on the individual process
steps of their department and because they were
already involved in the Late Show project via earlier
process tests. Based on the recordings and notes, the
interviews were transcribed and encoded. In order to
present the current Late Shipment process and the
design of the new Late Show process, the data
collected for both designs is merged and processed
into Swimlane diagrams using Lucidchart tooling.
According to White (2004), Swimlanes have the
ability to represent different functional
responsibilities. Further, Swimlanes are part of the
Business Process Modeling Notation (BPMN)
technique, which scores best in the research of
Recker, Rosemann, Indulska, & Green (2009) on the
degree of completeness for the system with the
features of the one analyzed in this work.
Figure 2: Current Late Shipment process diagram.
An Analysis and Design for the Repair Process of Late Show Shipments in the Export Cargo Process at SPL HUB
125
3 RESULTS
In this section, first is described the Current process.
Next, the Data Analysis is presented. After that, the
New process designed is shown. Finally, the
comparison of the results obtained to the objective is
explained.
3.1 Current Process
In the acceptance process of KLM Cargo at SPL, the
FOH message is generated after the shipment is
unloaded from the truck and certain acceptance
checks are conducted. In case the shipment does not
require any further checks, a ride is generated by the
Warehouse Management System (WMS) for the
shipment from the receiving area to either the storage
facility or the buildup buffer (B&B). At the buildup
buffer, the shipments are made ready for
transportation. The moment the shipment is delivered
at the right location, the RCS message will be sent,
which means the shipment is accepted and considered
ready for carriage. This process can be seen in Figure
1.
It is found that this standard acceptance process is
also followed when the shipment is delivered late.
Thus, the late shipment is always accepted (RCS
message). Further, it is experienced that currently late
shipments are often still transported to the buildup
buffer, while the WMS gives the instruction to drive
late shipments to the storage. The Flight Planners are
responsible for regularly checking the status of
shipments and whether the shipments on their
respective flights are already delivered to the freight
building. However, in practice it is seen that,
especially when these flights are departing close after
each other, the Planner does not have time to
continuously check this for each flight. The Planner
will screen at least three hours before the flight
departs, which is the moment that the buildup buffer
closes, whether the shipments are at that location.
Because the late shipments are often still transported
to the buildup buffer and because the Planner does not
always immediately know that a shipment is
delivered late, it will often be tried to build up the
shipment, even though there is sometimes not much
time left. Besides, due to a fault in the WMS, in case
a Planner decides to offload a shipment from the
planned flight because it is late, the shipment can still
be scanned and build up by the buildup employees.
Because of this, the Planners are reluctant with
rebooking and offloading shipments, and it is often
tried to build up a late shipment on the planned flight.
However, in case it was eventually not possible for
the buildup employees to build up the late shipment
on the planned flight, the shipment is left on the
buildup buffer and must be rebooked. This process
can be seen in Figure 2.
3.2 Data Analysis
From the statistical analysis on the total number of
shipments and the number of Late Shows, it is seen
that there are on average 282 shipments and 11 Late
Figure 3: Boxplot number of Late Shows per day for each day of the week (January, February & March combined).
ICORES 2022 - 11th International Conference on Operations Research and Enterprise Systems
126
Shows delivered per day in the combined first three
months of 2021. Figure 3 and Tables 1 and 2 show
that this number is highest on Monday, Tuesday, and
Friday with 14 Late Shows. Compared to the other
days of the week, the number of Late Shows per day
on Sunday is relatively low, with on average 3 Late
Shows. Further, on Tuesday, Friday and Saturday are
the largest deviations in the number of Late Shows
per day with a standard deviation of 7 or 8 Late
Shows. In the distribution of the total number of
shipments per day of the week, it is seen as well, that
Friday and Sunday are the busiest and least busy day
respectively. Further, on Friday and Saturday, the
number of total shipments is hard to predict, as the
standard deviation is largest on these days as well.
Continuing with the more detailed analyses on the
characteristics of the Late Shows, it is calculated that
73% of the Late Shows are between 1 and 120
minutes late and almost 49% of the Late Shows have
a delay between 1 and 60 minutes. Applying the
80/20 rule, the data analysis shows that 80% of the
Late Shows have a delay between 1 and 152 minutes.
Next, a categorization of the Late Shows based on
the product type shows that the top 3 products are
responsible for more than 67% of the total number of
Late Shows. Individually, product types 1, 2 and 3
represent 30%, 26% and 11% respectively. Another
product type follows shortly after product type 3
based on the number of Late Shows. However, this
product is generally delivered to freight building 1,
while this research is focused on freight building 3.
The eleven other product types represent small
numbers of Late Shows compared to the already
mentioned products.
Table 1: Data on number of Late Shows per day for each
day of the week.
Day
Data points
Total observations
M
13
180
T 13 178
W 13 152
Th 12 138
F 13 179
Sa 13 130
Su 13 38
Table 2: Statistics on number of Late Shows per day for
each day of the week.
Day
Mean
Med
Min
Max
Std Dev
M
13.9
14
6
24
5.9
T 13.7 11 1 31 8.1
W 11.7 12 5 19 4.3
Th 11.5 10 5 21 4.7
F 13.8 11 2 31 7.6
Sa 10 7 1 25 7.2
Su 2.9 3 1 6 1.5
Finally, an analysis on the DEP-R shows that 13%
of the Late Shows were rebooked after the RCS
Figure 4: General acceptance process of KLM Cargo at SPL including adjustments for Late Show process.
An Analysis and Design for the Repair Process of Late Show Shipments in the Export Cargo Process at SPL HUB
127
message was sent, indicating that these shipments did
not depart on the planned flight at RCS. As explained
in the introduction, the customer expectations are not
met for these Late Shows, which deteriorates the
quality of the process. It is assumed that this
percentage is caused by the late delivery of the
forwarder.
3.3 New Process
Figure 4 presents the adjusted acceptance process
considering the new Late Show process. This process
is applicable for all product types. The red highlighted
objects in figure 4 present the changes of the process
compared to the diagram in figure 1. As can be seen
in figure 4, it is checked in this new process whether
the shipment is late, after the FOH message is
generated. In case the shipment is late, different sub-
processes follow based on the type of product.
Because of the different characteristics of each
product type, the Late Show process is not applicable
to all products. Currently, the earlier mentioned
product types 1 and 3 are considered for the Late
Show process because of their high number of Late
Shows and thus the large operational impact. In
addition, they have a similar operational process in
the warehouse. The product with the second largest
number of Late Shows (product type 2), has a
different operational process in the warehouse,
because of certain storage requirements, which means
this product type requires a different Late Show
process. The same applies to the other products.
Besides, many of the other product types represent
small numbers of Late Shows, which means the high
effort it requires to include these products in the Late
Show process results in a marginal effect on the
operation.
The Late Show process designed in this research
is specifically for product types 1 and 3. These
product types follow the extra steps presented in
figure 5 because of the Late Show process. For
product types 1 and 3, a message is sent to the WMS
indicating the shipment is late. Based on this
message, the WMS automatically blocks the
shipment with the Late Show block. This block
triggers the standard Late Show process, which is
designed in figure 6.
Because of the Late Show block, the Late Show
will not be driven to the buildup buffer or the storage,
as a ride is generated in the WMS to the FOH-buffer.
This FOH-buffer is currently located in the receiving
area of the warehouse. With the Late Show in the
FOH-buffer, time is taken for the decision-making
process. The Planner checks for the Late Show in
correspondence with the operation on whether the
planned flight is still operationally achievable in the
remaining time before flight departure. In case the
Planner decides this is possible, he removes the Late
Show block in the WMS, which generates a new ride
from the FOH-buffer to the buildup buffer or the
storage facility. In case the Planner decides it is not
desirable to let the Late Show depart on the planned
flight, the Late Show first has to be rebooked, after
which the Late Show block is removed, and the
shipment is transported to the buildup buffer or the
storage facility. When the Late Show arrives at this
location, the RCS message is triggered.
3.4 Objectives
As mentioned in the introduction, there are two
objectives set by KLM Cargo for this project. The
design of the Late Show process should result in a
DEP-R improvement of 0.5% and 98% of the
shipments should be delivered on time at the buffer
for buildup.
Starting with the first objective, considering the
Late Show process designed in this research, the Late
Shows should no longer be rebooked after the RCS
message is sent. This results in an improvement of the
DEP-R. Considering the earlier mentioned 13% of the
Late Shows, the DEP-R should be improved with
0.51%. This expected improvement meets the
objective of 0.5% almost exactly. However, this
Figure 5: Late Show process steps in acceptance process (product types 1 & 3).
ICORES 2022 - 11th International Conference on Operations Research and Enterprise Systems
128
result assumes all product types, while it was
concluded earlier that the Late Show process is not
applicable to all products. When only the product
types 1 and 3 are considered, the DEP-R is improved
with 0.33%. In this case, the objective is not met, but
there is still a significant improvement. Finally, a
calculation can be made which includes product type
2, as this product represents a large part of the total
number of Late Shows. With the product types 1, 2
and 3, the DEP-R is improved with 0.37%. This
means that the objective of a DEP-R improvement of
0.5% is still not met.
Regarding the second objective, the Late Shows
will only be transported to the buildup buffer in the
designed Late Show process when it is possible to
build up the shipment for the planned flight. Thus,
when the buildup buffer is closed, the Late Shows
will not be delivered to the buffer. This results in an
improvement of the percentage of shipments that are
on time at the buildup buffer. It is calculated that in
all earlier described scenarios, the objective to have
98% of the shipments on time at the buildup buffer is
met. Thus, this objective is already met with the Late
Show process designed in this research (only
considering product type 1 and 3), with an expected
improvement of 0.38%.
3.5 Limitations
During the course of the research, there have been
discovered a couple of bugs and other differences
between the two data sources used for the data
analysis. It was found that there were significant
differences between the numbers of Late Shows
reported by the two sources. Therefore, it was hard to
make a good validation of the results of the data
analysis in this research. It was decided to continue
the data analysis using the appointed leading data
source, instead of waiting for the two sources to show
similar values. Therefore, the results of this research
could deviate slightly after the identified issues have
been solved and a proper comparison between the two
sources is made. However, the data analysis in this
research does provide an indication that is useful for
future decisions concerning the Late Show process.
Figure 6: Late Show process diagram (for product types 1 & 3).
An Analysis and Design for the Repair Process of Late Show Shipments in the Export Cargo Process at SPL HUB
129
The second data source is used as back-up in case
information was not available in the leading data
source. Further, it must be considered that the data
analysis does not make a distinction between freight
building 1 and 3, as this is not possible in the leading
data source. However, the designed process is
focused on freight building 3.
4 CONCLUSIONS
This research found that without a Late Show process,
there is no difference in the physical handling of a
shipment in the acceptance process of KLM Cargo in
case the shipment is delivered late. This means that
the shipment is always accepted, and the Received
from Shipper (RCS) message is always sent,
indicating the carrier considers the shipment as ready
for carriage. As a result, it is often tried to still build
up a late shipment and let it depart on the planned
flight, even though there is sometimes not much time
left for this process. This research shows that 13% of
the Late Shows do not depart on the planned flight at
RCS meaning the commitment to the customer is not
met, which deteriorates the quality of the process.
Based on this research, it is concluded that the
design of a standard Late Show process must consider
specific Late Show characteristics. It is important to
consider the product types requiring a different design
because of specific operational processes in the
warehouse. Further, the design of the process should
include a check on whether the shipment was
delivered on time, before acceptance of the shipment
and thus the sending of the RCS message. Then, in
case the shipment is late, it should be checked
whether the planned flight is still operationally
achievable. When this is not possible, the Late Show
first has to be rebooked to another flight before the
RCS message is sent. By following this decision-
making process while the shipment is stored in the
FOH-buffer located in the receiving area of the
warehouse, and thus before the shipment is driven to
the buildup buffer, it is assured that the Late Show
will be on time at the buildup buffer for the booked
flight.
The research recommends to also include at least
the product with the second largest number of Late
Shows in the Late Show policy, as this product also
has a large operational impact caused by the Late
Shows. This means further research must be
conducted into mainly the storage requirements of
this product, and how a Late Show process for this
product should look like. Further, it is recommended
to further research the predictability of the Late
Shows in more detail, as this research shows the
number of Late Shows is currently difficult to predict.
A more detailed analysis can lead to even better
insights into the characteristics of the Late Shows and
can help to decide on future improvements for the
process. In addition, it is recommended to validate the
data from the two data sources used in this research,
after the identified issues that caused the differences
between the two sources are solved
ACKNOWLEDGEMENTS
This research is conducted in the context of a
graduation internship for the study Aviation
Operations at the Amsterdam University of Applied
Sciences at the request of internship company KLM
Cargo.
REFERENCES
Blonk, M. (2017). Dock & Yard Management at KLM
Cargo. Master Thesis. Eindhoven University of
Technology. The Netherlands.
Henriksson, F. T., & Petersson, J. (2019). Mapping of the
Air Freight Handling at Stockholm Arlanda Airport.
Bachelors Thesis. Linköping University.
IATA. (2019). IATA Master Operating Plan Glossary.
IATA/ Cargo iQ. Retrieved March 8, 2021
Recker, J., Rosemann, M., Indulska, M., & Green, P.
(2009). Business Process Modeling- A Comparative
Analysis. Journal of the Association for Information
Systems, 10(4), 333-363. doi:10.17705/1jais.00193
Slack, N., Brandon-Jones, A., & Johnston, R. (2016).
Operations Management (8th ed.). Pearson.
White, S. A. (2004, July 6). Introduction to BPMN. Retrieved
from BPTrends: https://www.bptrends.com/bpt/wp-
content/publicationfiles/07-04%20WP%20Intro%20to%
20BPMN%20-%20White.pdf
ICORES 2022 - 11th International Conference on Operations Research and Enterprise Systems
130