The world’s leading publication for data science, AI, and ML professionals.

Bandung City Public Transportation: Service Coverage Analysis

How is Bandung City's public transportation service coverage? Has every region been served? A geographical analysis.

Street Network of Nearby Bandung (Source: Author, 2020; ©OpenStreetMap Contributors)
Street Network of Nearby Bandung (Source: Author, 2020; ©OpenStreetMap Contributors)

Hands-on Tutorials

How is Bandung City’s public transportation service coverage? Which regions are yet to be served by the public transportation network in Bandung City?

Introduction

Bandung City is the capital of West Java Province in Indonesia and it has its own transportation sector problem. At one time, it is considered to be one of the most congested cities in Indonesia, and the 14th in Asia (ADB, 2019). This is shocking because it is expected that Jakarta to be the most congested city, as Jakarta is the capital city and also is popular for its congestion, but Jakarta is second (for instance, Coconuts Jakarta (2019), Detik (2019), and many more). Bandung City is even worse than Mumbai. This makes me, a resident of Bandung City, think; what makes the congestion so bad? How do we solve the congestion problem?

Bandung City Spatio-Temporal Jamming Factor for 1 Day (Source: Author's Analysis, 2020; Basemap: ©OpenStreetMap Contributors; © CARTO licensed under the Creative Commons Attribution-ShareAlike 2.0 license (CC BY-SA 2.0))
Bandung City Spatio-Temporal Jamming Factor for 1 Day (Source: Author’s Analysis, 2020; Basemap: ©OpenStreetMap Contributors; © CARTO licensed under the Creative Commons Attribution-ShareAlike 2.0 license (CC BY-SA 2.0))

Well, the inevitable solution is for citizens to walk, cycle, or not to travel at all. While walking and cycling are limited to the traveler’s stamina, Public Transportation becomes the solution. Public transportation enables long-distance travel as it uses petroleum/another source of energy thus saving travelers’ stamina. With shared spaces, public transportation makes transportation energy and spatially efficient and these are why it is promoted.

In this essay, public transportation becomes the topic of discourse. Access to public transportation should be easy and logical for citizens to use. As stated, this access relies on human stamina, and with limited stamina comes limited travel distance. Thus, the distance to public transportation should be as close as possible to the citizens. I bet that you won’t take public transportation for a 7 km walk, but what if it’s just a 300 m walk? Cutting the distance short ensures low energy cost that is spent by citizens to walk or cycle to a public transportation stop. In my opinion, This is the first priority in public transportation planning. Hence a question emerges in my mind: How is the distribution of distance to public transportation? and which area of Bandung city has the furthest distance to public transportation?

With knowing the area that is far from the public Transportation, we can plan or propose a new route that provides public transportation service. By knowing the distribution of distance we can understand the existing condition of public transportation accessibility. This distribution can then be compared to other cities or even can be analyzed along with population data for further correlation analysis, but that’s another topic for another essay. Let’s just focus on the simpler questions first.

This essay is divided into 6 sections.

  • First, Introduction: you’re reading right now. this section explains the background of this essay;
  • Second, Objectives and Method: the goal this essay trying to achieve and the way this essay achieves;
  • Third, About Bandung City: where Bandung is, its relationship with Jakarta and the general description of Bandung City’s public transportation;
  • Fourth, Analysis: conducting the analysis;
  • fifth, Conclusion: answering the main question and concluding the result;
  • Sixth, Recommendation: what to do next to solve the problem and do improvements.

Objectives and Method

The objectives of this essay are to describe the distance to the nearest public transportation network in Bandung City, and the identification of the furthest area from the public transportation network (geographically).

The main geographical technique of analysis is Euclidean Distance analysis from the public transportation network. This Euclidean Distance’s result is in form of continuous spatial data (raster) and thus can be treated statistically as interval/ratio data. The difference from conventional/typical statistical data is that it has a spatial property thus we can make a map about it.

The description of the distance will be done by understanding the distribution of the distance yielded from the euclidean distance. This will be done using histogram analysis and we do not need the spatial information.

The geographical area identification will be visualized and analyzed using a map. Raster data is practically an image, so we can make a stretch of color to visualize the values. To make things concrete and practical for bureaucracy and policy planning, I’ll take a zonal statistic based on the smallest administrative boundary unit (Kelurahan). This is done because the nature of the interpretation with the raster that is produced for this analysis is fuzzy (Robinson (2009), Zadeh (1988)). Although I must say, I like doing it fuzzy as it explicitly shows what it is, but to make a policy we have to make the boundaries clear.

Additionally, the aforementioned identification (using a map) will be done by conducting Anselin Local Moran’s I (clustering) analysis. This analysis is conducted to identify regions that are strategic to be intervened. The analysis takes geographical relationship into account, thus spatial relationship and causality can be inferred. Don’t worry, I won’t bore you with the math. The point of this analysis is just to identify clusters.

Essay Framework (Source: Author, 2020)
Essay Framework (Source: Author, 2020)

About Bandung City

for this section, I’m going to divide into 2 subsections. The first subsection is Jakarta and Bandung; this explains a brief geographical characteristic and relationship between Bandung and Jakarta. The second subsection is Bandung City Public Transportation; this explains brief information about Bandung’s public transportation.

Jakarta and Bandung

Map of **** Jakarta and Bandung Built-Up Environment in Java Island, False Coloring of Landsat Image (Source: Author's analysis, 2019; contains altered content from _Landsat-8 image courtesy of the U.S. Geological Surve_y)
Map of **** Jakarta and Bandung Built-Up Environment in Java Island, False Coloring of Landsat Image (Source: Author’s analysis, 2019; contains altered content from _Landsat-8 image courtesy of the U.S. Geological Surve_y)

Bandung City is located on Java Island, Indonesia. It serves as the capital of West Java Province. The city is the core of the Bandung Metropolitan Area which includes West Bandung Regency, Bandung Regency, and Cimahi City. As the figure illustrates above, Bandung is not as wide as Jakarta. Bandung has a relationship with Jakarta and has been intensifying for some years. The development of Cipularang Toll Road connecting Jakarta and Bandung even brought more traffic and investment for Bandung City (Dorodjatoen (2009)). This Cipiularang Toll Road is illustrated by the bold line in the map below; the road is located in the southern part of the city.

Map of ** Bandung City Road Network 1:275.000 Scale** (Source: Author's Analysis, 2020; ©OpenStreetMap Contributors license (CC BY-SA 2.0))
Map of ** Bandung City Road Network 1:275.000 Scale** (Source: Author’s Analysis, 2020; ©OpenStreetMap Contributors license (CC BY-SA 2.0))

Bandung City Public Transportation

There are 2 main types of the city’s public transportation that the citizens can use to get anywhere (within the city): Bus network and Minibus network. The minibus network is similar to the Jeepney in the Philipines. This minibus is called "angkot" and it is ubiquitous. The bus network is just a typical bus that can be found anywhere in the world. The bus is operated by a state-owned enterprise Damri Company. Although between cities public transportation exists and often are operated by private companies, it is not within this essay’s topic. The network is illustrated in the map below.

Map of **** Bandung Metropolitan Area Public Transportation Network and Bandung City Administrative Boundary (Source: Author, 2020; ©OpenStreetMap Contributors license (CC BY-SA 2.0); route data: angkot.web.id)
Map of **** Bandung Metropolitan Area Public Transportation Network and Bandung City Administrative Boundary (Source: Author, 2020; ©OpenStreetMap Contributors license (CC BY-SA 2.0); route data: angkot.web.id)

One main function of the system in Bandung is that empirically there are no public transportation stop points. You can embark anywhere, and disembark anywhere as long as you are embarking/disembarking within the line. There are public transportation stops but it’s not institutionalized within the society.

Another means of public transportation are the taxi and ride-hailing vehicles that are facilitated by Gojek and Grab (just like Uber). Ride-hailing has its own issue because The Government’s policy does not recognize ride-hailing as public transportation. Nevertheless, these non-route base public transportation means are not the issue in this topic.

Analysis

This section is divided into two subsections: distribution of the distance, and geographical identification of the furthest distance.

Distribution of Distance

Using Euclidean Distance, a raster of 10-meter resolution is yielded. The Y-Axis explains about the frequency of the cells. So, the Y-Axis graph below represents the real area if we multiply it by the 100-square meter. As it illustrates, the average distance is 387.4 meters. This is great because the ministry of spatial plan guidance states that the Transit-Oriented Development area is delineated 800 m from a transportation facility. Although, we must note that 387,4 m is euclidean and not based on network analysis. This means that we assume that people fly straight to the transportation network, so this analysis is more of an approximation. Nevertheless, there are still regions that have more than 800 m distance.

Histogram Graph of the Distance to the Nearest Transportation Network (Source: Author's Analysis, 2020)
Histogram Graph of the Distance to the Nearest Transportation Network (Source: Author’s Analysis, 2020)

Geographical Identification— Euclidean Distance

Now, this map below is the result of the Euclidean Distance analysis. It is a 10-meter resolution raster image that contains numbers that are the distance to the nearest transportation network. A darker cell represents a further distance to the nearest blue line (public transportation network). The furthest distance is 3918.99 meters.

Fuzzy Analysis Using Euclidean Distance (Source: Author's analysis, 2020; route data: angkot.web.id)
Fuzzy Analysis Using Euclidean Distance (Source: Author’s analysis, 2020; route data: angkot.web.id)

The raw euclidean distance raster is fuzzy. There is no boundary to determine which area is high or low. Of course, we can compare it to the aforementioned 800 m TOD distance, but it truncates the statistical properties of the far distance. For example, this treats regions with 900 m distance the same with 3,000 km distance. To me, a more rational approach is to identify clustering. Identifying clusters helps because we can make new routes to the clusters. This is rational because by adding one line of the network to one of the clusters (which is supposed to be one region), it impacts the regions around the cluster. This effectively shorts the distance.

Anselin Moran’s-I is a spatial statistic method that is used to identify clusters by understanding the spatial relationship of the data. The result is a classification for each geographical unit and there are 4 classes: High-High, High-Low, Low-High, and Low-Low. High-high means that the region’s value(or geographical unit) is surrounded by high-value regions. High-Low means that the region’s value is surrounded by low-value regions, and so on. The low/high of one geographical unit is determined statistically by comparing it to the other units.

In this case, there are 3 identified clusters: High-High, Low-Low, and High-Low as the map below illustrate. There are 8 High-High regions, and there is 1 High-Low region. A Low-Low region means that the distance surrounding the region is short which is good and right now it is not in this essay’s interest. A High-Low region is a region where it has a long distance to the network, but it is surrounded by regions that are close. Now, there are many High-High regions in the eastern part of the City which I will say: these (eastern) regions have the worst accessibility to public transportation in Bandung City.

Distance Cluster Analysis (Source: Author's Analysis, 2020; route data: angkot.web.id)
Distance Cluster Analysis (Source: Author’s Analysis, 2020; route data: angkot.web.id)

Conclusion

Bandung City is considered to be the most congested city in Indonesia (ADB, 2019). For that, public transportation inevitably must be promoted. One of the ways is by cutting the distance to public transportation short. This way, the accessibility of public transportation can be improved as it is easier for people to walk. This essay explores Bandung City’s distance to public transportation and the main method that this essay uses is by conducting Euclidean Distance analysis.

First, this essay analyzes the distribution of the distance from the Euclidean Distance. The average distance of the regions is 387.4 meters. This is great because it is lower than 800 m walking distance suggested by The TOD Guidance by the Ministry of Spatial Planning. However, the histogram shows that there are regions that are more than 800 m far from the nearest public transportation.

Second, this essay produces 2 maps: The Euclidean Distance map, which is a fuzzy analysis; and the Cluster map (using Zonal Statistics and Anselin Moran’s I to the lowest administrative boundary (Kelurahan)) to support policy planning. Most of the regions that are far from public transportation are located in the eastern part of the City. These eastern regions are strategic for a new public transportation plan to be made.

Recommendation

Now we know that the eastern part of the City’s accessibility to public transportation is bad. In the future, this problem needs to be addressed by planning new routes. The local must be consulted to better plan the route destinations.

This research only identifies distance but there are regions that do not require public transportation. For example, the airport runway as runway takes a lot of lands. This research identifies an airport runway as a region that is far from the public transportation network, so we need to take into account the land use behind the distance. Perhaps, there are reasons for these regions to be inaccessible. The next research should take land use (or land coverage) into account.

Lastly, the next analysis needs to take population into account. Just like land use, perhaps, some regions just don’t have residents. Even though the regions are far from the public transportation network, it is not logical to plan a transportation route to a deserted place where no one will travel to or travel from in a huge capacity (to ensure business feasibility). Except, there are some region development plans taking into account.


Related Articles