Article citation information:
Passarella,
R., Nurmaini, S., Rachmatullah,
M.N., Arsalan, O., Kurniati, R., Aditya, A., Afriansyah,
I.G., Fathan, R., Yousnaidi, R.S., Veny, H. Data analysis of commercial
aircraft landing on the runway airports in Indonesia. Scientific Journal of Silesian
University of Technology. Series Transport. 2023, 120, 233-247. ISSN: 0209-3324. DOI: https://doi.org/10.20858/sjsutst.2023.120.15.
Rossi PASSARELLA[1], Siti NURMAINI[2],
Muhammad Naufal RACHMATULLAH[3], Osvari ARSALAN[4], Rizki KURNIATI[5],
Aditya ADITYA[6],
Indra Gifari AFRIANSYAH[7], Rifqi FATHAN[8],
Rani Silvani YOUSNAIDI[9],
Harumi VENY[10]
DATA ANALYSIS OF COMMERCIAL AIRCRAFT LANDING ON THE RUNWAY AIRPORTS
IN INDONESIA
Summary. A runway
excursion is a runway safety issue in which an aircraft exits the runway
improperly. Therefore, we believe it is necessary to conduct a data
investigation of commercial airplanes landing at the airport in order to provide
criticism, learning, and social control to airlines and air service
policymakers in Indonesia so that they can become involved in protecting public
transportation users, particularly those using air transportation. This study
focuses on the procedural safety of commercial airplanes landing on the runway
at airports in Indonesia. The data utilized for landing analysis is historical
ADS-B data collected and acquired from the Flightradar24
website in 2019. The approach utilized in this study is exploratory data
analysis (EDA), in which we explored the dataset to find trusted data (128
tier-1 data points). The results of this study showed that the aircraft that
was the object of research was unlikely to land at the Touch Down
Zone (TDZ), while the rest showed that its landing
was highly likely outside the TDZ.
Keywords: ADS-B,
procedural landing safety, runway safety issue, touch down zone
1.
INTRODUCTION
Whether an aircraft is flying in conformity
with safety rules is not always known to passengers on commercial flights. The
one thing passengers know is that the plane landed safely, with no turbulence
and a smooth landing. Whereas, conceptually, airplanes have safety procedures
in all phases of flight [1,2]. Therefore, this study
was carried out to comprehend the issue of flight safety and inform passengers
(users of air transportation) about the landing phase. It is necessary to
convey and inform this idea so that pilots and co-pilots would take precautions
and propriety when performing landings in accordance with the procedure to
reduce incidents and accidents that could result in losses for all parties.
Indonesian air transportation experienced rapid
growth starting in 2000 with the entry of a new trend, namely Low-cost carrier
(LCC). This is due to population factors, economic
growth, and the reach of the archipelago, causing the air transportation
industry to be an option for the transportation of people and goods. However,
with this growth, the number of accidents has also increased, caused by runway
excursions, including runway overrun. A runway excursion is a runway safety
issue in which an aircraft exits the runway improperly. Runway overruns happen
when an aircraft is unable to stop before reaching the end of the runway.
Runway excursions can occur as a result of pilot error, bad weather, or an
aircraft malfunction [3].
Therefore, we believe it is necessary to
conduct a data investigation of commercial airplanes landing at the airport in
order to provide criticism, learning, and social control to airlines and air
service policymakers in Indonesia so that they can become involved in
protecting public transportation users, particularly those using air
transportation. This study focuses on the safe landing of a commercial
aircraft on the runway at Indonesian airports. With the advancement of aircraft
monitoring utilizing ADS-B (Automatic Dependent Surveillance-Broadcast)
technology, numerous global researchers have been able to further investigate
information and research on commercial aviation in order to improve flight
safety [4-6]. The ADS-B data exploration is an important lesson in improving
the development of aircraft navigation systems with the advantages offered by
this system, such as easily accessible data, and real-time flight operational
data. Thus, this technology can help provide increased commercial aviation
security.
This study focuses on ADS-B investigation data
on the safety procedures of commercial aircraft landing on runways at airports
in Indonesia. In order to determine whether the aircraft landed in the safety
zone and, if it did not, whether the speed was within the acceptable range. In
addition, the research subject in the form of a commercial airplane in this
paper does not mention its registration code (PK-???)
but is only mentioned as an airplane.
The paper is organized as follows: the
background section contains information regarding an accident or incident that
occurred on an Indonesian airline as a result of an overrun during the landing
phase. In order to have better knowledge of a safe aircraft landing, subsequent
studies are reviewed, including a literature review of the theory supporting
the landing strategy. The data and methodology parts provide the information
and the methods to reach the goal. This section also discusses the study's
limitations. The results and discussion section cover the findings and outcomes
of the data. Finally, in the conclusion section, we highlight the contributions
of this work.
2.
BACKGROUND
Several aircraft mishaps in
Indonesia, including runway overruns, have also raised concerns about
commercial aviation safety and comfort. According to the National
Transportation Safety Committee (Komite Nasional Keselamatan Transportasi-KNKT) report and analysis [7], there were 69
occurrences on Indonesian territory between January 2015 and August 2022,
according to data on overrun accidents as shown in Fig 1. It was noted to have
a peak incidence of 14 occurrences in 2016. The trend then turned downward
until 2019, but then turned upward again in 2020, reaching 11 events and
representing an increase of 83.3%. Papua province is the biggest contributor to
the record of overrun events, contributing 28 times, or 41%. This data also
shows that the province of Papua requires special attention to the transportation
industry and air flight safety [8]. Furthermore, based on data analysis of
accident data given by the official KNKT agency from
1997 to 2020, pilots and co-pilots (human factors) contributed to a significant
number of accidents and incidents in the history of commercial aviation in
Indonesia, amounting to 52.6% [9].
Fig. 1. Summary of KNKT reports based on aircraft accidents and incidents due
to overrun from 2015-2022
Another study [10], revealed that
the total aviation accidents in Indonesia from 2007 to 2014 recorded 228
aircraft accidents, with 80 of them being RE (Runway Excursion), or aircraft
accidents attributable to RE accounted for 35% of the total accidents during
the 8-year period.
In the meantime, according to a data
analysis and statistical overview of commercial jet aviation accidents issued
by Boeing in 2021, fatal accidents in the RE category occurred 7 times between
2012 and 2021 and were the second-greatest occurrence after the loss of
control-in-flight (LOC-I)[11]. Meanwhile, the rate of fatal incidents reported
from 2012 to 2021 was 50% when examined in the flight phase (Fig. 2). This
proportion indicates the high likelihood of landing mishaps.
Another comparable research included only overrun
instances that happened between 1980 and 1998 in Australia, Canada, the United
Kingdom, and the United States [12]. 137 (76%) of the 180 commercial aircraft
occurrences happened during landing, while 43 (24%) occurred during take-off. ACRP also investigated overrun and undershoot incidents to
construct a global database [13].
According to the FAA, in 90% of overrun incidents, the
aircraft exceeds the runway end at 70 knots or less and
mostly stops within 306 meters of the extended runway centerline.
Therefore, 306 m long RESA is recommended by FAA to
provide enough braking space for aircraft with 70-knot speed [14,15].
Therefore, in order to reduce the number of overrun
events, it is necessary to investigate the runway of an airport in Indonesia.
During the observation process, data is needed about the location or position
where the aircraft is declared touchdown. These statistics can be obtained from
ADS-B data. In addition, the rules regarding runways and landing zones that
have been established are also referred to in this research. So that the
conclusions obtained are facts.
Based on [16], in his presentation, the touch-down
zone or point (TDZ) is 1/3 of the runway length,
which is marked with several markings on the runway (Fig. 3). In the event that
a commercial aircraft is unable to land at TDZ, the
aircraft should turn and attempt another landing. Additionally, according to
studies cited on the website of the flight safety organization, 52% of airport
aircraft landings involve steady approaches, while 48% involve unstable
approaches. But in reality, this landing approach is affected by several
variables, one of which is TDZ, with the main landing
gear on both sides of the runway centerline. Go around if all the main landing
gear is located on one side of the centerline. The next variable is that if
there are still 2,000 feet (609.6 meters) of runway to cover, then the aircraft
speed must be less than or equal to 80 knots.
As a result, we can use these 2 variable points as a
guide for conducting research on commercial airplanes.
Fig. 3. The airplane touches
down zone (TDZ) [16]
3.
DATA AND METHOD
The approach taken to answer the research objectives
is how to get data, how to process data, and how to generate data insight.
ADS-B is being used to provide an alternative to the
use of conventional radar in commercial flight monitoring. ADS-B data is public
and can be viewed and monitored through several sites, one of which is flightradar24. The B737 MAX 8 was
one of the aircraft whose data we were interested in, and we tracked its
journey since the tragic aviation accident by retrieving all of its flight
history data until the day the Indonesian aviation authorities decided to
ground it due to the ongoing investigation into the B737
MAX 8 crash.
The Boeing 737 MAX 8 airplane with registration in
Indonesia territory (production year 2017) that executes phase landings at
airports in Indonesia was the subject of this investigation. The data utilized
for landing analysis is historical ADS-B data collected and acquired from the flightradar24 website in 2019.
The flight recorded data is taken from March 19, 2018,
to October 2, 2019, which is equivalent to 1 year, 6 months, and 14 days (582
days). According to the raw data, during this duration, the aircraft made a
total of 1400 landings, with 1227 of them being national landings (Indonesian
jurisdiction) at 33 airports, while the rest were landings at overseas
airports. The distribution map of airports landed is shown in Fig. 4, with the
largest bubble data point being Ngurah Rai
International Airport, Denpasar (IATA: DPS- with 105 landings), followed by Hassanudin International Airport, Makassar (UPG- with 60 landings) and Supadio
International Airport, Pontianak (PNK- with 38
landings).
Fig. 4. Bubble distribution
map of airport data points visited by
the aircraft object during its 582 days of commercial flight service
The national landings made by the aircraft for 582
days are represented by a proportion of 61% of the total number of landings
from the three airports (DPS, UPG, and PNK). Therefore, 3 airports that are frequently landed by
aircraft were analyzed and looked at for data insights and data patterns to
gain knowledge about compliance with landing procedures.
3.1. Data
In terms of procedure, the dataset was prepared in
multiple steps, beginning with a visit to the Flightradar24
website and then obtaining the aircraft-specific data saved for each flight in
*.csv (Comma Separated Values) format. 1400 flying observation data points were
collected. These data were then processed in a variety of ways, including
confirming each flight has its landing data, specifically the landing altitude
indicator with a value of 0. This technique reduces the data to 414 flight data
points; statistically, the data reduction at this stage is 70%, implying that
only 30% of the data is viable for the next step. Following the collection of
414 flight data points, the next stage is to separate data with the ICAO (International Civil Aviation
Organization)-recommended quality for ADS-B data, namely tier 1 (update
message interval less than 10 seconds), tier 2 (update message interval less
than 20 seconds), and tier 3 (update message interval greater than 20 seconds
and less than 60 seconds). The results of separating data based on this tier
revealed that tier 1 had up to 128 flight data points, tier 2 had up to 200
flights, and tier 3 had up to 86 flights.
This study used tier 1 data, as much as 128 flight
data, implying that only 9% of the data from the 1400 database was utilized.
Three airports were chosen from this 9% for analysis, which is referred to as a
dataset (Fig. 5). This dataset will be used in the following procedure to
achieve the study objectives. The method part will go over the specifics of
this explanation.
Fig.
5. The reduction data is based on the quality of the ADS-B update message
interval
3.2. Method
The approach utilized in this study is EDA
(Exploratory Data Analysis) in which we explored the dataset to find trusted
data (128 tier-1 data points) with 14 landing airports in 128 flight data
points. These 14 airports are then ranked to determine the three airports with
the highest frequency. Denpasar (DPS), Pontianak (PNK),
and Makassar (UPG) are the three target airports for
inquiry, having 66, 22, and 6 landings data, respectively.
Fig. 6. The
approach taken to analyze aircraft landings in this study
After determining the three airports, the next step is
to map the GPS position of the observed aircraft landing, which is based on
latitude and longitude data. Following the plotting with Google Maps, the next
step is to detect the TDZ. From the plotting and TDZ results, it is known how much flight data lands in the TDZ and how much is outside the TDZ.
Scatter plots and violin plots are then used to visualize the data. An
illustration of our approach is depicted in Fig. 6. While the tool used to
visualize data is Scimago Graphica
[17], for data processing, namely orange data mining [18].
3.3. Limitations
In addition to the investigative data material used
and the approach taken to obtain data insight, it is also necessary to explain
the limitations of the research we conducted. First, the data used is secondary
flight data taken from the ADS-B system stored on the filightradar24
website. This data has undergone several calibrations that are explained on its
website to ensure that the data presented and stored is reliable. Flightradar24 has 100% coverage in Europe and America, but
in the following countries: Canada, Mexico, the Caribbean, Venezuela, Colombia,
Ecuador, Peru, Brazil, South Africa, Russia, the Middle East, Pakistan, India,
China, Taiwan, Japan, Thailand, Malaysia, Indonesia, Australia, and New Zealand
are well covered, especially around airports [19].
The second step is to standardize data. In this study,
we track the performance of ADS-B data using the standard message update
interval, namely the maximum update interval (x<10) [20].Third, the data that is the
focus and is trusted to be studied is data that is included in tier-1(x<10)
such that data transfer before and during landing occurs in less than 10
seconds. Fourth, we only utilize one airplane as an example of a landing for
study purposes. Fifth, the study is limited to the top three airports where the
airplane lands.
4.
RESULTS AND DISCUSSION
Based on the research method approach, the results and
findings, as well as the data insights that can be known and learned from the
flight data stored on the flightradar24 (ADS-B) site,
are described in this section. Each target airport is presented with the
results and followed by a discussion.
In processing the dataset of 128 flight data, we found
that the aircraft we observed underwent a rough landing in the form of bouncing
during landing, which is indicated by altitude 0. Then the next data shows that
the altitude changes and returns to altitude 0. We assume this phenomenon is
known as a hard landing. The number of flights that experienced this was 5,
with the most frequently occurring airport being Denpasar, and the rest being
Pontianak. The identification of this hard landing data is shown in Tab 1.
Tab. 1
Indications that the aircraft experienced
a hard landing based on historical data
|
Callsign |
Date |
Airports |
|
LNI 3829 |
14/09/2018 |
Denpasar |
|
LNI 988 |
26/10/2018 |
Pontianak |
|
LNI 2610 |
04/11/2018 |
Denpasar |
|
LNI 2622 |
09/03/2019 |
Denpasar |
|
LNI 2622 |
12/03/2019 |
Denpasar |
4.1. Finding - Denpasar (DPS)
I Gusti Ngurah
Rai International Airport (IATA: DPS, ICAO: WADD), also known as Denpasar (DPS), is an international
airport in Tuban Village, Kuta
District, Badung Regency, about 13 kilometers from
Bali's capital, Denpasar. The airport has two runway directions, 09 and 27, and
a 3,000-meter-long asphalt runway surface. It is Indonesia's second-busiest
airport.
According to the data set, the number of aircraft
landings with ADS-B data quality according to ICAO
standards (Tier-1), namely update message intervals of less than 10 seconds, is
66, which is further divided into two headings or landing directions, namely
runways 09 and 27, with 44 and 22 landing data points, respectively.
Touch Down Zone (TDZ) Runway 27 Runway 09 Touch Down Zone (TDZ)
Fig.
7. Data points for the aircraft object landings at Denpasar Airport (DPS)
runways 09 and 27
The landing data points were then transformed into
data visualizations as shown in Fig 7. The visualization used is a scatter
plot. Scatter plots are commonly used to show the relationship between two
variables displayed on a graph. The x-axis shows the range of runway 27
starting from a zero point at the beginning of the runway, while the y-axis
shows the meter distance from the landing point of the aircraft based on ADS-B
data (latitude and longitude) measured from the base point of runway 27 using
equation (1).
To determine how far the aircraft’s ADS-B
transponder began to detect from the base point of runway 27, it is necessary
to calculate the distance of two GPS (Global Position System) location points
(latitude and longitude). This study’s calculations used the Haversine equation. This formula (equation 1) is used
because the earth’s surface is affected by a curvature [21].
Harvesine formula:
The symbol r is the earth’s radius (6371 km),
and the latitude and longitude are
Fig. 8(a) depicts the runway 27 landing data point,
where only 5 data points show the aircraft landing at the TDZ
point, while the remaining 17 points are outside the TDZ.
In other words, only 23% of landings that have ADS-B data that meets the
standards show the aircraft pilot landing in accordance with the safety zone.
Looking deeper into the landing data, the next
variable is the aircraft's speed when landing in two zones, inside and outside.
As a result, we employ the violin plot to properly visualize it (Fig 8(b)). A
violin plot uses density curves to represent numerical data distributions for
two groups (0 stands for Inside TDZ and 1 for outside
TDZ). Each curve's breadth correlates to the
estimated frequency of data points in each location. To offer extra
information, densities are frequently accompanied by an overlaid chart type,
such as a box plot.
LNI3829 LNI2622 LNI034 LNI3829 LNI2622 LNI034 LNI2630 LNI3829 LNI3829 LNI2630
Fig.
8. Results of the aircraft landing data visualization at
Denpasar airport (DPS) on runway 27
(a)
Data points are plotted in a scatterplot, based on their distance from
the threshold bar of runway 27.
(b)
Violin Plot based on aircraft speed versus abnormalities
(0 stands for Inside TDZ; 1 for Outside TDZ)
We can see two groups of landing zones in Fig. 8(b).
The violin plot in Group 0 shows that the distribution of aircraft speed data
during landing varies between 140-160 knots, but the distribution of speed data
in Group 1 shows that the aircraft landed outside the TDZ.
This demonstrates that even though the aircraft is landing inside two-thirds of
the runway length, the speed is already below 100 knots. However, the data
reveals that three data points indicate that the aircraft landed on two-thirds
of the runway with speeds over 100 knots. This is a point of concern for us.
Based on the violin plot, it can be seen that aircraft
landing outside the TDZ (between 1000 meters and 3000
meters) have an average speed below 80 knots, except for five landings
exceeding the recommended speed. By rule, these 5 landings should go around and
re-land because if they are still forced to land, there is a risk of overrun.
The data points of the five flights can be seen in Fig 8 (a) and (b).
The following analysis is of runway 09, which is still
at Denpasar airport. A scatter plot in Fig. 9(a) shows that 22 aircraft
landings are within the TDZ and 22 are outside the TDZ, indicating that 50% of the observed aircraft landing
data on runway 09 is within the TDZ and vice versa.
Meanwhile, if we look at Fig. 9 (B), namely the violin plot, it can be seen
that landing in the TDZ has an average aircraft speed
of 144 knots, and outside the TDZ in the zone between
1000 meters and 3000 meters on runway 09, the average speed of the landing
aircraft is 87 knots, of which 13 landings out of 22 outside the TDZ exceed the 80-knot procedure speed, or in other words,
59% of the total landings outside the TDZ should
perform the go-around procedure.
Thus, it can be concluded that of the 66 landings made
at Denpasar airport, 27 were in the TDZ while 39 were
outside the TDZ, with 18 of those outside the TDZ at a risk of overrun.
Denpasar airport (DPS) on runway 09
(a)
Data points are plotted in a scatterplot, based on their distance from
the threshold bar of runway 09
(b)
Violin Plot based on aircraft speed versus abnormalities
(0 stands for Inside TDZ; 1 for Outside TDZ)
4.2. Finding - Makasar
(UPG)
Makasar Airport (UPG), also known
as Hasanudin Airport, has four runways, which are
numbered 03, 13, 21, and 31. The combination of runway 03 and 21 has a length
of 10171 ft x 148 ft (3100
meters’ x 45 meters), while the next combination is runway 13 and 31,
with a length and width of 8202 ft x 148 ft (2500 meters x 45 meters). A surface of asphalt spans
these four runways. Based on the dataset, the airplane flight was detected to
have good data quality (tier-1) for only six landings out of the 60 landings it
had made during the database collection (582 days). This means that only 10% of
landings at the airport have good ADS-B data quality.
The six landings were made on runway 31, which has a
runway length of 2500 meters. In the landing history data, it was found that
only 2 flights landed at the TDZ point (33.3%), while
the remaining 4 landed outside the TDZ point. The
landing position at this airport based on GPS points from ADS-B data can be
seen in Fig. 10(a), while when viewed from the violin plot (Fig.10(b)), it can be seen
that the 2 landings in the TDZ have an average speed
of 158 knots, while the 4 landings outside the TDZ
zone have an average speed of 59.75 knots.
Fig.
10. Results of the aircraft landing data visualization at
Makasar Airport (UPG) on
runway 31
(a)
Data points are plotted in a scatterplot, based on their distance from
the threshold bar of runway 31.
(b)
Violin Plot based on aircraft speed versus abnormalities
(0 stands for Inside TDZ; 1 for Outside TDZ)
Furthermore, based on Fig 10(a) and (b), the airplane
was detected landing at a point of 1990 meters from the starting point of
runway 31, which means that there are only 510 meters left where the end of
runway 31 will end, despite the fact that the aircraft's landing speed is only
69 knots. However, what the pilot did on flight LNI795
(Departure from Jayapura) on July 23, 2018, has a risk that endangers
passengers and aircraft in the form of overrunning.
4.3. Finding - Pontianak (PNK)
According to Pontianak airport information, it has a
runway with two landing directions, namely runway 15 and runway 33. In the
filtered data, there were 8 B737 MAX aircraft that
landed at this airport and had tier 1 quality; 20 landings were found, all of
which landed via runway 15. For reference, runway 15 measures 7382 feet by 148
feet (2250 meters by 45 meters) and has an asphalt surface. The TDZ zone has a length of 2460 feet (750 meters) measured
from the beginning of runway 15.
According to Fig 5, only 20 of the 38 landings made by
the aircraft during its 582-day flight history have Tier-1 data quality. Or, if
converted into a percentage, the value is 52.63%. In the 20 flights of Tier-1
data, only 4 flights made landings at TDZ (20%); the
remaining 16 other flights made landing points outside TDZ.
Fig. 11(a) and (b) displays two landings, LNI686 and LNI988, on runway 15
of Pontianak airport (PNK), exceeding 2/3 of the
runway, leaving only 1/3 of the runway with speeds of 84 and 92 knots,
respectively. With a speed value of this magnitude and the remaining runway
being less than 750 meters (1/3 the length of the runway), the aircraft must
perform a maximum speed reduction, which can affect the lifetime of the braking
device. This maximum braking action was forced by the pilot to avoid
overrunning.
LNI988 24 Nov 2018 LNI686 LNI686 LNI988 24 Nov
2018 LNI988 26 October 2018 LNI988 26 October 2018
Pontianak airport (PNK) on runway 15
(a)
Data points are plotted in a scatterplot, based on their distance from
the threshold bar of runway 15.
(b)
Violin Plot based on aircraft speed versus abnormalities
(0 stands for Inside TDZ; 1 for Outside TDZ)
4.4. Summary of findings
Based on the findings of the landing data at the three
airports, the aircraft that became the object of research showed that it was
unlikely to land at the TDZ, while the rest showed
that it was highly likely to land outside the TDZ,
which should be a safety procedure to take go-around action. For more details,
see Tab. 2.
Tab. 2
Summary of findings
|
Airport |
Tier 1 ADS-B Data |
TDZ |
Non TDZ |
|
Denpasar (DPS) |
66 |
41% |
59% with 46% Risk of Overrun |
|
Pontianak (PNK) |
20 |
20% |
80% with 12.5% Risk of Overrun |
|
Makasar (UPG) |
6 |
33% |
67% with 25% Risk of Overrun |
5.
CONCLUSION
The investigation into two B737
MAX 8 incidents that occurred in late 2018 and early 2019 led to the decision
to halt this aircraft as of October 3, 2019, according to a study of landing
history data for this aircraft obtained from March 19, 2018, to October 2,
2019. We discovered that, according to the ICAO-established
message interval quality standard, only 9% of the ADS-B data from the flight
history of the analyzed aircraft can be used for analysis. We solely
concentrate on the top three airports, Denpasar (DPS), Pontianak (PNK), and Makassar (UPG).
Additionally, we discovered that there were five rough landings, with Denpasar
Airport (DPS) accounting for most of them (Tab. 1). Furthermore, data analysts
showed that this aircraft fleet is unlikely to land in TDZ
zones and does not comply with go-around procedures if the landing point is
outside the TDZ (Tab. 2).
The results of this study can be utilized to help
airlines and air service policymakers in Indonesia evaluate current practices
and acquire additional data, so they can actively defend passengers using
public transportation, especially those traveling by air.
Acknowledgment
We would like to thank everyone who
contributed to the data collection process (data engineering).
Credit author statement
Rossi
Passarella: conceptualization, methodology, investigation, writing-original
draft preparation. Siti Nurmaini:
supervision, reviewing data curation, validation. Muhammad Naufal
Rachmatullah: visualization, investigation, software.
Osvari Arsalan: formal
analysis. Rizki Kurniati:
formal analysis. Aditya Aditya: investigation. Indra Gifari Afriansyah:
investigation. M Rifqi Fathan:
investigation, software. Rani Silvani Yousnaidi: investigation. Harumi Veny:
data curation, reviewing and editing, formal analysis.
References
1.
FAA.
„Airplane flying handook (FAA-H-8083-3c)(Flight Standards Service).
Available at:
https://www.faa.gov/regulations_policies/handbooks_manuals/aviation/airplane_ handbook.
2.
FAA.
„Airline Safety”. Available at: https://www.faa.gov/other_visit/aviation_industry/
airline_operators/airline_safety.
3.
FAA.
„Runway excursion”. Available at:
https://www.faa.gov/airports/runway_safety/ excursion.
4.
Busyairah S.A. 2016. „System specifications
for developing an Automatic Dependent Surveillance-Broadcast (ADS-B) monitoring
system”. International Journal of Critical Infrastructure Protection
15: 40-46. ISSN: 1874-5482. DOI: https://doi.org/10.1016/j.ijcip.2016.06.004.
5.
Lambelho
M., M. Mitici, S. Pickup, A. Marsden. 2020. „Assessing strategic flight
schedules at an airport using machine learning-based flight delay and
cancellation predictions”. Journal of Air Transport Management 82:
101737. ISSN: 0969-6997. DOI: https://doi.org/10.1016/j.jairtraman.2019.101737.
6.
Schultz
M., J. Rosenow, X. Olive. 2022. „Data-driven airport management enabled
by operational milestones derived from ADS-B messages”. Journal of Air
Transport Management 99: 102164. ISSN: 0969-6997. DOI: https://doi.org/10.1016/j.jairtraman.2021.102164.
7.
KNKT.
„Laporan Investigasi”. Available at: https://knkt.go.id/investigasi.
8.
Lucyana
N. 2022. „Implementasi metode k-means dalam analisa penyebab kecelakaan
pesawat di indonesia”. Undegraduate
Thesis. Universitas Sriwijaya.
9.
Oktariani.
G. 2022. „Analisis Kecelakaan Penerbangan Di Indonesia Menggunakan Metode
Agglomerative Hierarchical Clustering”. Undegraduate Thesis. Universitas Sriwijaya.
10.
Saputra
A.D. 2017. „Studi analisis penyebab runway excursion di indonesia
berdasarkan data komite nasional keselamatan transportasi (KNKT) tahun
2007-2016”. Warta
Ardhia 43(2). ISSN: 0215-9066. DOI:
https://doi.org/10.25104/wa.v43i2.305.93-104.
11.
Boeing.
„Statistical summary of commerciat jet airplane Accidents: Worldwide
Operations 1959-2020”. Available at: https://www.boeing.com/resources/ boeingdotcom/company/about_bca/pdf/statsum.pdf.
12.
Kirkland
I.D.L., R.E. Caves, M. Hirst, D.E. Pitfield 2003. „The normalisation of
aircraft overrun accident data”. Journal of Air Transport Management
9(6): 333-341. ISSN: 0969-6997. DOI:
https://doi.org/10.1016/S0969-6997(03)00033-4.
13.
Shirazi
H. “ACRP Report 3: Analysis of Aircraft Overruns and Undershoots for
Runway Safety Areas”. National Academies Press. Available at: https://repository.lboro.ac.uk/
articles/report/ACRP_Report_3_Analysis_of_aircraft_overruns_and_undershoots_for_runway_safety_areas/9461387.
14.
FAA.
„AC 150/5220-22B: Engineered Materials Arresting Systems (EMAS) for
Aircraft Overruns”. Available at: https://www.faa.gov/documentLibrary/media/Advisory_Circular/
AC_150_5220-22B.pdf.
15.
Global
Runway Safety Simposium. „Engineered arresting systems corporation
emasmax Aircraft Arresting Systems for Runway Overrun Protection”.
Available at: https://www.icao.int/safety/RunwaySafety/GRSS2011/Documentation/Docs/ESCO
Zodiac Aerospace Solutions Workshop Presentation.pdf.
16.
Stephens.
A.T., M.H. Smith. 2012. „Anatomy of an Overrun- what the data tell us
about why overruns occur”. In: International Air Safety Seminar.
Santiago, Chile. Available at:
http://www.fof.aero/presentations/F/o/F/2/0/1/3/Tuesday/013_2013_Boeing.pdf.
17.
SCImago.
(n.d.). „VIZ TOOLS-SCIMAGO GRAPHICA”. Available at:
https://www.scimagojr.com/viztools.php.
18.
Demšar.
J., T. Curk, A. Erjavec, Č. Gorup, T. Hočevar, M. Milutinovič,
M. Možina, M. Polajnar, M. Toplak, A. Starič, M. Štajdohar,
L. Umek, L. Žagar, J. Žbontar, M. Žitnik, B. Zupan. 2013.
„Orange: Data Mining Toolbox in Python”. Journal of Machine
Learning Research 14: 2349-2353. ISSN: 1532-4435. Available at:
http://jmlr.org/papers/v14/demsar13a.html.
19.
Flightradar24.
„How flight tracking works”. Available at :
https://www.flightradar24.com/ how-it-works.
20.
ICAO
ASIA and Pacific Office. „ADS-B Implementation and Operations Guidance Document.
Edition 14.0”. Available at: https://www.icao.int/APAC/Documents/ edocs/Revised%20ADS-B%20Implementation%20and%20Operations%20Guidance%
20Document%20(AIGD)%20Edition%2014.pdf.
21. Robusto C.C. 1957. “The
Cosine-Haversine Formula”. The American Mathematical Monthly 64(1):
38-40. ISSN: 0002-9890. DOI: https://doi.org/10.2307/2309088.
Received 18.12.2022; accepted in revised form 29.02.2023
Scientific Journal of Silesian University of Technology. Series Transport is
licensed under a Creative Commons Attribution 4.0 International License
[1] Department of Computer
Engineering, Faculty of Computer Science, Universitas
Sriwijaya, Palembang, Indonesia. Email:
passarella.rossi@unsri.ac.id. ORCID ID
:https://orcid.org/0000-0002-7243-0451
[2] Intelligent System Research Group.
Faculty of Computer Science, Universitas Sriwijaya, Palembang, Indonesia. Email:
sitinurmaini@gmail.com. ORCID ID:
https://orcid.org/0000-0002-8024-2952
[3] Department of Informatics, Faculty of Computer Science. Universitas
Sriwijaya, Palembang, Indonesia. Email: naufalrachmatullah@gmail.com. ORCID ID:
https://orcid.org/0000-0003-3553-3475
[4] Department of Informatics, Faculty of Computer Science. Universitas
Sriwijaya, Palembang, Indonesia. Email: osvari.arsalan@ilkom.unsri.ac.id. ORCID
ID: https://orcid.org/0000-0003-3474-5812
[5] Department of Informatics, Faculty of Computer Science. Universitas
Sriwijaya, Palembang, Indonesia. Email: rizkikurniati@ilkom.unsri.ac.id. ORCID ID: https://orcid.org/0000-0002-4737-2659
[6] Department of Informatics, Faculty of Computer Science. Universitas
Sriwijaya, Palembang, Indonesia. Email: aditya728391@gmail.com. ORCID ID: https://orcid.org/0000-0002-5950-0638
[7] Department of Informatics, Faculty of Computer Science. Universitas
Sriwijaya, Palembang, Indonesia. Email: indragif@gmail.com. ORCID ID: https://orcid.org/0000-0002-9763-491X
[8] Department of Informatics, Faculty of Computer Science. Universitas
Sriwijaya, Palembang, Indonesia. Email: muhammadrifqifathan@gmail.com. ORCID ID: https://orcid.org/0000-0002-0854-6214
[9] Department of Informatics, Faculty of Computer Science. Universitas
Sriwijaya, Palembang, Indonesia. Email: ranisyach@gmail.com.
ORCID ID:
https://orcid.org/0000-0002-6435-7012
[10] College of Engineering, Universiti
Teknologi MARA. Shah Alam
40450. Malaysia. Email: harumi2244@uitm.edu.my. ORCID ID: https://orcid.org/0000-0001-6604-3554