Study of Integration of Greater Bandung BRT Service in Bandung City

The Trans Metro Pasundan (BRT) requires a comprehensive review for network integration with existing modes like angkot, Damri Bus, Trans Metro Bandung (TMB), and commuter trains (KRD). This study aims to employ a GIS model to assess BRT integration in Bandung City through spatial and trip analysis. Key network integration indicators include: (1) overlap level; (2) area coverage within catchment areas; (3) the number of transfers between corridors; and (4) route availability. Analysis shows that the unclear intermodal hierarchy has caused high overlap level and low coverage areas. Low intermodal connectivity also leads to high number of transfers. These issues do not meet the concept of equity set out in the regulations. BRT overlap exceeds 60% with angkot and 90% with Damri, failing to enhance coverage in certain areas, particularly four eastward sub-districts, with catchment ratios of just 13% - 27%. Utilizing ArcGIS, 26 out of 900 trips between sub-districts, originating at BRT stops, require more than the regulation’s designated two corridor transfers. Route availability is measured by the walking distance needed to access a stop between routes, which remains below the 500-meter regulatory limit.


Introduction
The BRT Trans Metro Pasundan, a public mass transport program developed by the Ministry of Transportation, GIZ, and local governments, will be integrated with other modes of transportation to reduce congestion in the area.The BRT program in Greater Bandung is planned to start operating in stages until 2023.Trans Metro Pasundan is present as a BRT transportation to support community mobilization which is integrated with other mass transportation services.In addition, it is proposed that the BRT development will use direct services to be more flexible and have a wider coverage, so that passengers are not forced to transfer.
Currently, the Bandung City Transportation Agency is also designing an integrated transportation system that is assisted by the British Embassy through the Future Cities program, where there will be an evaluation of the public transportation system that has been operating in Bandung which is divided into feeder, main, and branch routes.The main routes will maximize bus modes, namely BRT, Trans Metro Bandung, Damri.The public transport modes that are currently operating in Bandung City consist of Trans Metro Bandung (TMB), Damri, commuter train (KRD), and angkot, within the city and across cities. Branch networks can be filled by angkot because it must be integrated with mass transportation.The integration of existing public transportation service areas in Bandung City as of 2020 has been analysed, namely angkot, Damri Bus, Trans Metro Bandung, and KRD by Putra [1].Several areas in Bandung City are still not covered by the existing public transport catchment areas, namely in the eastern part of Bandung in Mandalajati, Ujung Berung, Cibiru, Panyileukan; to the middle areas, namely Arcamanik and Cinambo; the southern part, namely Gedebage and Rancasari; and the southwest, namely Babakan Ciparay and Bojongloa Kidul.Therefore, it is necessary to review whether BRT services can provide additional coverage in areas in Bandung City that have not been served well.
In this sense, the main objective of this study is: "To conduct a study of integration of Greater Bandung BRT Service in Bandung City" to answer the research question: "Is Greater Bandung BRT service integrated in Bandung City?".The sub-objectives are to identify the catchment areas of public transport services in Bandung City, to develop GIS models for evaluation of the integration of public transport services, and to evaluate the new public transport services in terms of catchment areas and integration (case study: proposed BRT Trans Metro Pasundan).Due to limited data and conditions, the scope of this study is limited to Bandung City only.The modes studied are angkutan kota (angkot), Damri Bus, Trans Metro Bandung, KRD (Bandung local train), and BRT Trans Metro Pasundan.The transit routes are also based on existing legal documents.This research focuses on network integration, from the spatial and trip perspective.The analysis of this study was primarily conducted in 2021, hence the data of public transport network routes is using the 2021 version, while the O-D projection data available and used in this study is for 2031.

Literature review
One of the indicators that is most often developed in studies on the integration between modes of public transport is the overlap level.A study conducted on the optimization of integration between metro and bus modes mentions that competition and cooperation between the metro and public transportation are very important in an integrated public transportation network [2].Another study also uses an indicator of the overlap level in assessing public transportation connectivity, which can be quantified by the ratio of overlapping route lengths [3].The overlap analysis between routes can also be based on the coverage area along the route where it can trigger a reduction in passengers and competition between fleets [4].
Another research on intermodal integration conducted in other cities analysed integration with the identification of catchment areas at railbus stops, as well as a potentially multimodal travel approach from determining the origin to destination zone (O-D matrix) [5].Analysis of the characteristics and patterns of user travel demand, characteristics of origin-destination within and outside the city, is also studied, using a quantitative approach, namely spatial analysis (overlap) [6].An analysis related to network or route integration consisted of an analysis of route overlap and the availability of routes that accommodate direct trips between nodes, along with an analysis of the number of mode changes [7].
Regarding spatial analysis in the study of public transport, a study evaluated the relationship of the spatial distribution of the population to the public transportation network to ensure that the population's needs are met, with components of bus coverage area, population density, and origindestination (O-D) data [8].GIS was utilized as a mapping in evaluating the effectiveness and coverage of public transport areas.The developed indicator or parameter system is spatially correlated with the public transportation network and population distribution [9].Another research that applies GIS in spatial analysis is the modelling of the accessibility of public facilities with an analysis of the coverage area of public transport supported by data in the form of population density and land use in public facilities, especially residential areas [10].
IOP Publishing doi:10.1088/1755-1315/1294/1/0120213 However, there are still several aspects in the evaluation of the public transport network that have not been investigated further.In the technical study of the Trans Metro Pasundan BRT network planning conducted by Bandung City Transportation Agency [11], the actual overlap level has been studied, but only between BRT corridors, not yet on modes in Bandung City that have the potential to become a competition, including angkot.Research that evaluates public transport networks based on population density in the region is not commonly found in Indonesia, even though the population is a component that needs to be supported by public transportation to ensure accessibility [10].Analysis of the availability of feeder routes and the hierarchy of public transport routes is the minimum service standard stipulated in the guidelines in Indonesia, namely Minister of Transportation No. 10 of 2012 [12], the Decree of the Director General of Transportation No. 687 of 2002 [13], and Guidelines for Assessment of the Level of Integrated Intermodal Transportation [14].However, these factors have not been widely analysed in Indonesia.

Catchment area
One aspect that needs to be considered to conduct a public transit analysis is the catchment area, which is related to the potential users who are interested in public transit, depending on the location of bus stops and routes of public transit.The catchment area is an area that is still quite comfortable to walk to transit routes or stops [15].The size of the service area of a route is very dependent on how far the walking distance still feels comfortable.The walking distance tolerance of each location varies, depending on geographical conditions, climate, land use characteristics, walking habits, and walking goals [16].Analysis of service area instead of buffer area will lead to better results for spatial or service area analysis [17]. Figure 1 shows a comparison of the analysis of service areas and buffer areas.Analysis of the buffer area of the image seems to include more space, so more area is affordable.However, given the fact that people will not be able to reach the bus stop directly from a bird's-eye view, a better analysis is through the analysis of service areas by assuming people will use existing roads in the catchment area.

Indicators of public transit network evaluation
In Law No.22 of 2009 concerning Road Traffic and Transportation [18], it is explained that in the urban scope, the Regency/City Regional Government is obliged to ensure the availability of public transportation for the transportation of people and/or goods within the Regency/City area within its administrative area.From this statement, accessibility of public transportation is the aspect that is focused on in this case.
Various guidelines or previous studies indicate that the transfer or trip of modes or corridors in the public transport network should be minimized.A good public transportation route follows the trip pattern of transportation passengers to create a more efficient trip so that the mode of transfer that occurs when passengers travel by public transport can be minimized [13].Likewise, locations that have the potential to become travel destinations are sought to become service priorities.Public transportation is required to ensure the ease of passenger trips between corridors from origin to destination to ensure accessibility.The number of corridor transfers is set to a maximum of 2 (two) times [12].
The public transit integration model consists of the main mode and the feeder mode.The main characteristic of multimodal networks is to have a network that is connected between modes and recognizes the different levels of the network [19].The main modes are designed to provide services to residential areas, trade, and activity centres as well as transportation nodes.This planned service is on the main corridors to ensure high travel speed and reliability of operation.For Bandung City, the modes of public transit with greater capacity are the railway, the BRT, Damri Bus and Trans Metro Bandung, making it suitable to focus on becoming the main modes.Angkot that has a lower capacity, can be focused on operating in the feeder route, by reaching users in smaller areas and roads, then feeding it to the main modes.Angkot routes can be rerouted or restructured to support their function as feeder modes for city buses.The criteria for the integration of public transport service networks, both intermodal and the same mode, include the availability of public transport routes that serve trips between transportation nodes [14].In addition, the hierarchy of the route network that serves the trip is also reviewed, whether it is the main route or the feeder route.
Competition and cooperation between the metro and public transportation are very important in an integrated public transportation network [2].Therefore, the integration of the public transport network can be achieved if there is minimum competition between modes, namely between BRT and other modes.Bus modes (BRT, Damri, and TMB) and train modes (KRD) as the main modes should be supported by angkot as feeder mode.If the BRT route still overlaps with routes from other modes, they would compete over travel demands on the road, and integration of the public transport network will not be achieved.In terms of network efficiency and operation, route overlap should also be minimized.If the city transportation route is diverted to other areas or areas, that can cover a wider area so that it can feed passengers for city buses, the public transportation network can be said to be better integrated.In addition, a wider area coverage certainly increases the distribution of public transport areas as well.
The main factor in determining the network of public transport routes is that the existing public transport routes are sought as close as possible to reach areas with high potential demand, namely those with high population density [13].Area with a high intensity of activity has high trip generation and attraction, so it should be accessible by public transportation.However, apart from potential service areas, public transportation should also cover all existing urban areas.This follows the concept of equal distribution of services to the provision of public transportation facilities.The more public transit routes are distributed in a city, the more roads are reachable by the catchment area of public transit.

Methodology
The data used are secondary data from official documents issued by the Regional Development Planning Agency (Bappeda) in Bandung City, and various literatures that can support research.The proposed BRT will be reviewed against four other modes of development plans in Bandung City, namely angkot, Damri buses, TMB, and local trains or KRD.Therefore, information on the corridors of each mode is needed, accompanied by the location of the bus stops/terminals/stations scattered throughout Bandung City.The network and region data are superimposed or stacked into a map via ArcGIS.The O-D matrix between sub-districts is collected from the Department of Transportation of Bandung City to identify patterns of passenger trips in Bandung City.Sub-district density data to review public transport network services in meeting demand in Bandung City is obtained from Statistics Indonesia (BPS) 2020.

Trip analysis with Network Analyst
The integration of the BRT network with other modes is analysed by reviewing the trip of users.To analyse the trip taken in a trip in the transportation network system, including the number of transfers and the availability of routes in the public transport network, a modelling is needed that describes the network (link) and stops (nodes) containing information in the form of distance, time, obstacles, direction, and other attributes.The Network Analyst feature in ArcGIS can help find the best route (quickest path), create an O-D cost matrix, and create a model for route analysis, to help analyse the trip in Bandung City with public transportation to obtain predetermined integration indicators, namely the number of mode transfers and the route availability.Network Analyst on ArcGIS software version 10.3 is used.
A network dataset containing public transportation networks along with points or stops is built and equipped with information on distance, speed, direction, and route number.The public transport routes that pass-through Bandung City are 39 city angkot routes, 80 intercity angkot routes across Bandung City, 11 Damri Bus corridors, 6 Bandung Trans Metro corridors, and 1 KRD corridor, accompanied by plans for 12 BRT corridors.To avoid overlap in the model, the network dataset is made in three dimensions (3D) so that each route has a different elevation.The stops on each route are then connected to pseudo stops on the road network that have zero elevation, with connectors at each stop.The connector will also be used in the number of transfer attributes.After that, the start/end locations of the trip to issue the quickest route are put at the 30 BRT stops in the sub-district in Bandung City, or angkot/bus stops if there are no BRT stops available.The point is placed at one of the pseudo-stops on the road network, then the Network Analyst will decide to move through the connector to get to a stop on a route that is already at a different elevation or start the journey on foot on the road network to get to a stop on another route.With the user-based indicator method, the following two trip indicators can be identified.1.The number of intermodal transfers to public transport corridors that must be carried out by users when travelling to a destination in Bandung City.

Availability of routes in serving trips in Bandung City
The results of the analysis of the number of transfers and the route availability are then recapitulated in an O-D matrix to review the trips that have not been served properly, as well as to evaluate the Bandung City public transportation network.

Spatial analysis
The service catchment area is within walking distance of the route or transit stop.The total catchment area of each mode (angkot/city bus/train) can be calculated using ArcGIS software.Meanwhile, the overlap level between two public transport routes can use the following ratio [3].
where r1 and r2 are the notation for each route, olr1,r2 is the length of the overlapping route between the two routes; and lr is the total length of the route.In this study, the ratio was based on catchment area.Overlap analysis is performed per BRT route on each of the other modes' routes to see the proportion of bus routes that overlap with other modes.On the other hand, if a city bus route overlaps with many angkot routes, but the length of each angkot route is low, then the bus route is not interrupted by those angkot routes.
To review the distribution of public transportation routes in Bandung City in responding to demand, the ratio of the catchment area of public transportation available in each sub-district is reviewed on the population density indicating residential, and on the generation and attraction numbers related to trip and destination centres in Bandung City.The catchment area in each subdistrict is calculated based on the road [20], while the area that represents the public transportation catchment area is the total catchment area which has considered overlap.It is analysed whether areas with high density or becoming centres of departure and destinations are priorities for public transport services, especially with the presence of BRT in Bandung City.
Catchment area ratio in sub-district = Catchment area of public transport Catchment area of network (2) The performance of the integration of BRT network services with other modes, namely Damri buses, TMB, KRD local trains, and angkot in Bandung will be discussed further.The analysis of each indicator that is presented in maps/tables/graphics, namely the number of transfers, route availability, overlap level, and catchment area ratio, is reviewed to identify routes or locations for BRT stops or other modes that still do not provide good connectivity and answer the needs in a certain area.Then, recommendations are developed based on the indicators in this study, which include the proposed addition or rerouting of public transport routes and the addition of bus stops in certain areas.For public transit modes that will be reviewed, based on Bandung Mayor Decree in 2008, officially there are 39 angkot routes within Bandung City.There are also 80 intercity angkot routes across Bandung City.There are 11 Damri bus routes, with five of them connecting points within the administrative boundaries of Bandung City, while the remaining six connect across regions outside of Bandung City, namely Jatinangor, Tanjung Sari, and West Bandung.In addition to Damri, there are currently five Trans Metro Bandung (TMB) bus routes in the city.There is also the local Greater Bandung train or KRD, a rail-based mass transportation currently available in Bandung, which is a series of commuter local trains.With a length of 51.7 km, KRD serves stations from Ciroyom, Bandung Station, Andir, Cimindi, Cikudapateuh, Gedebage, and Kiaracondong.
Five regencies/cities in Greater Bandung are part of the development of road-based public transportation (BRT), according to the Minister of Transportation Regulation No. 9 of 2020.The BRT program in Greater Bandung under the name Trans Metro Pasundan is planned to start operating in stages until 2023.BRT will be integrated with other modes of transportation to reduce congestion in the area.Trans Metro Pasundan is present as a BRT to support community mobilization which is integrated with other mass transportation services.It is also proposed that BRT development will use direct service services to be more flexible and have a wider coverage so that passengers are not forced to change buses.12 corridors that will be developed in the Greater Bandung City that connect Bandung City with the surrounding area and stretch for 23 km.BRT connects points within Bandung City as well as with surrounding areas such as Cimahi City, Baleendah and Soreang in Bandung Regency, and Jatinangor in Sumedang Regency.Several areas in Bandung City which are the centres of departures and arrivals for BRT routes are Cicaheum Terminal, BEC Mall, and Dipatiukur.In conducting a spatial analysis of public transportation routes in Bandung, the radius of the catchment area is determined in advance based on the walking distance.The catchment area of public transportation is measured from transfer points, namely stops/terminals, because passengers are allowed to get on and off only at stops, such as bus modes (BRT, Damri, and TMB) and KRD.However, users of angkot in Bandung City go up and down the road, so the catchment area is assumed along the road.Therefore, the catchment area of an angkot is calculated from the roads traversed by each route.Assuming the catchment area of angkot is linear along the route, and its role as a feeder, it is different from bus and rail modes which act as the main mode which is larger and has a catchment area drawn from the bus stop, so the radius or walking distance for angkot modes is also different.Therefore, considering the characteristics of Bandung and the trip of its citizens, a walking distance of 200 m is used for public transit in the city, while for trains is 400 m. 400 m to city bus stops (BRT, Damri, and TMB), as well as stations that are passed by local trains or KRD.The road network also uses a 400 m radius [16].
Figure 7 shows the catchment area of each mode depicted through ArcGIS.The entire catchment area of the BRT (coloured orange) is covered by the catchment area of other modes.This phenomenon can be an early indication that BRT is not yet integrated with other modes because it serves the same area in Bandung City and does not provide additional accessible coverage areas.

Catchment area ratio
Table 1.Catchment area ratio compared to population density and traffic generation and attraction in Bandung City.
To evaluate the distribution of public transportation routes in the sub-districts of Bandung City, the ratio of the available public transportation service areas in Bandung City to the catchment area of the road network in each sub-district is calculated, as shown in equation ( 2) in Chapter III.The catchment area ratio shows the area in the sub-district that can be reached by public transportation.The ratio can be seen in Table 1, which shows the colour gradation for each sub-district against other sub-districts in that mode.Almost the entire road network in Sumur Bandung and Bandung Wetan sub-districts is served by public transportation, which is 98.8% and 92.3% respectively of the catchment area of the road network.When combined with other modes (the "Total" column is red), the entire road network in the two sub-districts is served by public transport (100%).For BRT mode, the catchment area ratio is highest in Andir and Sumur Bandung Subdistricts, which are busy areas in the centre of Bandung City.Overall, the sub-districts that have the lowest catchment area coverage of public transport are Cibiru (13.6%),Gedebage (21.0%), and Ujungberung (23.2%), an area in the eastern part of Bandung City.This means that more than 75% of the road network in the sub-districts is not served by any mode of public transportation.
If the catchment area ratio of public transportation is reviewed further on population density and the value of generation and attraction in each sub-district in Bandung City, it can be analysed how far the fulfilment of the needs of public transportation trips can be.The population density figure indicates residence, while the generation and attraction rates are related to the trips and destination centres in Bandung City.
The population density value in each sub-district is symbolized by a green circle in Figure 8, while the ratio of the total catchment area is indicated by the background colour in each sub-district.The worse the service area ratio, the darker the colour, which indicates a critical condition.Bojongloa Kaler sub-district as the most populous sub-district has a good catchment area ratio, i.e., 88% of the road network is covered by public transportation (large circles with light background colours).Therefore, the value of population density covered is still high, only decreasing from 39.57 thousand/km 2 to 34.88 thousand/km 2 .Other sub-districts that have a high population density and good public transport coverage are Andir and Astana Anyar.The sub-districts that have a relatively high population density but have a low catchment area ratio are Kiaracondong, Mandalajati, and Ujungberung Districtsindicated by the red boxwhich are in the north-to-east part of Bandung City.This does not meet the regional integration criteria that public transport routes are sought to reach areas with high demand potential, namely those with high population density [13].This phenomenon is related to the next spatial indicator, namely the overlap level in the area, namely BRT for angkot which reaches 60% and for Damri Buses with 92%.
On the other hand, the less densely populated sub-districts (small circles) tend to have a low catchment area ratio (dark background), as shown by the blue box in Figure 8, namely Cibiru and Gedebage sub-districts in the east and Cidadap in the north.This phenomenon shows that the service area of public transportation in Bandung City is uneven and does not meet the criteria [13] so it can hinder regional development.One of the goals of BRT mode planning is to reduce the gap in infrastructure services that can hinder economic growth, but this is not achieved because public transportation remains difficult to access for some people in Bandung City in some areas, even BRT does not provide additional accessible areas, as shown in Figure 7.  Apart from population or residential density, the catchment ratio of public transport areas in Bandung City is also reviewed against the generation and attraction values in each sub-district.Table 1 shows the comparison of generation and attraction currents in each sub-district to the catchment area ratio.The sub-districts that have good public transportation coverage are those with high traffic demand, such as Bandung Wetan, Batununggal, Coblong, and Regol, especially Regol which is the location of various terminals and stations in Bandung City.However, some areas have a relatively low public transport catchment area ratio compared to the generation and/or attraction demands, as shown by the yellow boxes in Figure 9 and Figure 10, namely Kiaracondong and Rancasari which have high generation demand for public transport, and Panyileukan for the attraction value.

Overlap level
The first overlap level analysis is to compare the BRT mode as the main mode to angkot as the feeder mode in Bandung City.The overlap level can be related to the previous spatial indicator, the catchment area ratio, as shown in Figure 11.Catchment areas that are darker in colour indicate a higher overlap level between the two routes.The red box captions indicate BRT routes in the northeastern part of Bandung City (Mandalajati, Ujungberung, and Antapani, Cibiru), and Rancasari in the south, which have a high overlap level with angkot.This condition reduces the coverage in the area, which can be seen in the dark colour background of the sub-district, as well as explains the previous phenomenon that the sub-district has a relatively high population density but a low catchment area ratio.Both BRT and angkot serve major roads, thus creating competition.This condition reduces coverage in the area because local roads are underserved, whereas angkot should instead cover small roads to increase the coverage area that is accessible and reduce overlap.Apart from angkot, the BRT corridors have even worse overlap level with other bus modes, namely Damri and TMB.And like angkot, the BRT corridor that overlaps with Damri is in the northeastern area, accompanied by corridors to and from Cimahi.The unclear division of the role of BRT with other main modes has resulted in high regional competition in some areas and corridors.Meanwhile, shown in Figure 13, the overlap level of the main modes other than BRT, namely Damri and TMB buses, and KRD to angkot is lower than BRT to angkot, where no corridor has an overlap level of more than 60 %, as happened in the BRT corridors.The phenomenon of BRT overlap level which is worse than the existing main modes is also shown by the graph of the percentage of route pairs that overlap at a certain level in Figure 14.The overlap level between routes dominates at 20% and below because the catchment area between routes must overlap at a low level to be connected and integrated with other modes of transportation nodes such as bus stops/stations.As many as 12.2% of the total pairs of BRT routes to angkot in Bandung City have an overlap level in the range of 20% to 40%, and when added with an overlap level of 40% and above, the value becomes 13.8% (12.2% + 1.2% + 0.4%).The overlap level for angkot across Bandung City at the level of 40% and above is also above this figure, to be precise at 15.6%.When compared with the percentage of pairs of existing main modes (Damri, TMB, and KRD) vs. urban angkot, the figures are not much different (13.0%), but there are no overlapping routes at the level of 60% and above.The percentage for overlapping levels of intercity angkot also decreased considerably to 8.7%.The conclusion is that the existing mode of overlap with angkot is better than the BRT overlap.The overlap level of BRT to the existing main modes (Damri, TMB, and KRD) is higher than that of angkot, which when added up is 28.7% at an overlap level of above 20%, indicated by grey bars.This phenomenon needs to be a concern to reduce competition among bus modes in Bandung City so that integration runs better.After the quickest route between sub-districts was issued, the use of the BRT mode was relatively minimal, namely only 150 out of 900 trips using the BRT mode more than 20% of the total distance traveled.This phenomenon can be caused by the BRT route which does not follow the trip pattern in Bandung City and has a high overlap level with the existing modes, so that the Damri, TMB, or angkot routes have been served, and the Network Analyst uses the existing modes.The first trip indicator analysed is the number of transfers between corridors, which are set a maximum of 2 (two) times in [15] and [16].However, based on the results of the analysis by Network Analyst for the combination of BRT, Damri, TMB, KRD, and angkot modes in Bandung, there are still 26 out of 900 trips (~3%) that require users to transfer three times, namely trip to Cibeunying Kaler and Sukasari sub-districts, as shown in Figure 16.This phenomenon can also be seen in the trip to and from the South Bandung area (blue box), which includes the areas of Bojongloa Kaler and Bojongloa Kidul, as well as Buah Batu.

Trip indicators
The next indicator in the integration of the public transport network is the route availability in a trip with public transport.To quantify the definition of route availability by Network Analyst, the distance to reach the stop was used, which was set at 300-500 m for the urban area and 500-1000 m for the rural area [13].In this study, the figure of 500 m is used to accommodate trips in the urban and rural areas.The distance to a stop is calculated using the walking attribute of the Network Analyst, which indicates the distance it takes to move from one corridor to another.Figure 17 shows a visualization of the walking distance indicator the user has travelled in the trip, and several trips require walking to transfer, but all of them are still under 500 m.Other trips are dominated by green lines which indicate the journey does not need to be taken on foot at all.Overall, users do not need to walk far to reach a stop to transfer in travelling in Bandung City by public transport, which means that at least the public transport routes are connected and available for other routes and meet the indicators of route availability.
26 inter-sub-district trips in Bandung City do not meet the indicators for the maximum number of transfers or route changes, which requires three transfers in one trip.The trip was dominated by the destination of Cibeunying Kaler.It was found that the 26 trips had a relatively low demand for transit, and even had no demand at all.Table 2 shows 20 inter-district trips with the highest transit demand in Bandung City.The trip only requires one transfer, some even do not need to transfer.The walking distance is also under 200 m, which means that every route taken is connected to other routes, so it can be said that public transportation in Bandung City meets the indicators of network integration in terms of the maximum number of transfers and the availability of routes for a trip with the highest flow.
However, the use of BRT in these trips should be noted, because only 2 of the 20 trips use BRT in their quickest routes, even though start/end points are located at BRT stops in each sub-district.The previous analysis of spatial indicators can explain this phenomenon, that the BRT overlap level is very high for angkot and bus modes, namely Damri and TMB, as well as Figure 7 that shows the catchment area of the BRT already served by other modes.In addition, the planned BRT route does not follow the trip pattern of users in Bandung City, which requires more travel time due to potential transfers if BRT is still used on the route by the Network Analyst.In other words, the existence of the BRT plan does not have a significant impact on Bandung City.From the results of the analysis, several issues were found related to efforts to integrate the public transport network in Bandung City, namely regional competition due to overlapping, low distribution of area coverage in certain areas, and the high number of transfers in several trips due to lack of intermodal connectivity, even after the BRT mode plan.On the other hand, the indicator of the availability of routes in trips is fulfilled.

Discussion: High BRT overlap level due to unclear intermodal hierarchy
The BRT corridors have a high overlap level, which has the potential to create competition with other modes.The Alun-Alun -Cibiru Corridor has the worst overlap level because the overlap is very high, both for angkot as feeders and for other main bus modes, namely Damri.The overlap level with the Alun-Alun -Jatinangor corridor via Antapani even reached 92.5%, completely replacing the role of the other corridor.These results indicate the lack of clarity in the hierarchy of the public transport network in Bandung, namely the role of BRT against other main modes, namely city buses, as well as for feeder transport, namely angkot, so that they fight over the passenger area.In addition, the Alun-Alun -Cibiru BRT corridor also overlaps with angkot, namely the A10 Cicaheum -Cileunyi route.The two corridors overlap along routes within Bandung City so that BRT has the potential to decrease the role of the angkot.Besides, Table 2 shows the minimal use of BRT mode in the quickest route with Network Analyst, for the trip with the highest demand for public transport use.This phenomenon can be further explained through Table 3 below, that some underused BRT corridors have a high degree of overlap with other main modes and/or angkot, as shown by the Elang -Cicaheum corridor.The corridor has an overlap rate of ~70% with the Damri Bus and TMB which also share the Cibeureum (Elang) -Cicaheum corridor.In addition to the two bus corridors, the Elang -Cicaheum BRT corridor also has an overlap rate of above 40% with many other corridors.This condition causes the BRT corridor to be used only 14 times in inter-sub-district trips in Bandung City for both directions, a very small number, as well as shows the existence of this corridor is less significant.
Other BRT corridors that are less used in the quickest routes are connecting Greater Bandung, such as Baleendah -BEC, Cimahi -BEC, and Soreang -BEC which ends at BEC, so it is natural that it is not used for trip within Bandung City, even though the overlap level with angkot and/or Damri buses is also very high.The Jatinangor -Dipatiukur BRT Corridor via the toll road is not being used at all because there is also a Damri Jatinangor -Dipatiukur bus corridor.This high overlap level harms public transport operations because they compete for passengers in the same service area.In addition, overlap causes headway irregularities or the time between fleet arrivals, where there are times with a busy number of fleets, but there are also hours that have irregular fleet arrival schedules.This is detrimental to public transport users.

Discussion: Lack of equitable coverage of public transport areas
In addition to causing regional competition between modes, the high overlap level also causes a decrease in regional coverage because they serve the same area.The role of the feeder, which should feed passengers to the main mode by serving smaller travel-generating areas that are not served by the main mode, is less visible due to the overlap.The issue of the lack of clarity in the hierarchy is again the cause of the lack of feeder roles in increasing the coverage area in Bandung City.Some areas in the BRT corridor that have high overlap with angkot have a low catchment area ratio, and this phenomenon is found from the results of spatial analysis.
The Alun-Alun -Cibubur BRT Corridor serves some of the same areas as the Gedebage -Simpang Dago angkot in the eastern area of Bandung City, a collection of sub-districts that have relatively high residential densities such as the Mandalajati, Ujung Berung, and Cibiru.It can be seen in Figure 18 that the area has a low catchment area ratio, which is in the range of 15% -25% only.This means that more than 75% of the road network in these sub-districts is not served by public transportation at all.There should be an opportunity for angkot to feed users from residential areas to use public transportation.One of the reasons for this low demand is the lack of public transportation in the area, making it difficult to develop the area if mobility to the area is not supported by mass transportation.This is also not following the concept of equity adopted by public transport, especially BRT which can be accessed by all classes of society to reduce the gap in infrastructure services that can hamper economic growth.This phenomenon also does not meet the concept of equal distribution of services to the provision of public transportation facilities that in addition to paying attention to potential service areas, public transportation also covers all existing urban areas [13].It is also stated that the availability of public transportation for people's transportation services must be guaranteed within the district/city area which is in the administrative area [18].

Discussion: Fulfilment of route availability indicator
An indicator of the availability of routes or other modes at the node for each trip is quantified by walking distance in the quickest route generated in the Network Analyst.From the analysis results visualized in Figure 17, all trips do not require a walking distance of more than 500 meters to reach a stop.Thus, every inter-sub-district trip in Bandung City has a route that can continue the journey from the previous route, and the integration of the public transportation network in Bandung City meets the indicators of route availability in the trip regulated in [12], [13], and [14].

Discussion: Number of transfers more than two due to lack of intermodal connectivity
Public transportation needs to minimize the number of transfers between corridors in trips in urban areas, as stipulated in the guidelines, in an effort to follow the pattern of trips in Bandung City.The number of corridor changes in urban areas is set at a maximum of 2 (two) times.Based on the quickest route shown in Table 2, indicators of the number of transfers and availability of routes have been met for trips with the highest demand for transit use, which indicates the public transportation network in Bandung City has followed the existing trip pattern.However, with the combination of BRT modes with angkot, Damri Bus, TMB, and KRD, the use of BRT for this trip is still very minimal.
In addition, there are still 26 trips in Bandung City that require 3 (three) transfers, especially towards Cibeunying Kaler from the south and east areas of Bandung City, which shows a lack of connectivity or intermodal integration in the Cibeunying Kaler area, especially towards BRT as a new mode plan that is expected to increase coverage in Bandung City.However, from another perspective, these 26 trips have a relatively low demand for transit usage or even no demand at all.The low demand for trip by public transport for these destinations can cause the trips to become less of a priority in planning public transport networks, including BRT and other modes.But on the other hand, public transport routes that are less integrated into facilitating the trip, on the contrary, can trigger low demand for the use of public transport.This does not meet the initial public transport planning objectives, such as one of the objectives of the BRT plan is to have a wider coverage so that passengers do not change buses.In the spatial and trip analysis, the BRT corridor plan has a high overlap level and low usage in the quickest route for trips between sub-districts in Bandung City.This indicates the lack of clarity in the hierarchy of the public transportation network in Bandung City, where BRT as the main mode should serve the main transportation corridor in Bandung City, which includes the trips from west to east in three areas in Bandung City, namely the north, centre and south, as well as the trips from the north (Bandung City) to the south (Soreang), as shown in Figure 19.There is also a need for the role of supporting corridors, both from BRT and other bus modes, to serve places that are the centre of trip in Bandung City, such as shopping centres, markets, and educational or office areas.Finally, angkot as feeder transport can fill gaps in areas that have not been served by main transport and supporting transport, as well as feed users to the corridor.
The following is a summary of the evaluation of network integration in Bandung City: • The overlap rate of BRT for angkot is 62%, and even 92% for Damri Buses.BRT dominates in overlapping levels above 40%, compared to overlapping existing modes vs. angkot.
• Due to the overlap, BRT does not increase coverage in some areas, such as 4 sub-districts in the eastern part of Bandung City, with a catchment area ratio of only 13% -27%.The role of angkot as a feeder is also insignificant, indicating the unclear hierarchy of the public transport network in Bandung.
• From the quickest route between sub-districts with start/end points at BRT stops, there are 26 out of 900 trips that require transfer between corridors more than 2 times [14].These trips dominate the destination of Cibeunying Kaler, which has low transit demand.

•
The indicator of route availability is met through quantification with walking distance to reach a stop to transfer, which is still below the maximum limit of 500 m [13].
Of the three issues of integration of the public transport network in Bandung City that have been discussed based on the analysis results, an example of scenario recommendations for each issue is compiled, namely removing overlapping BRT routes to reduce intermodal competition, rerouting angkot routes to reduce overlap while increasing regional coverage, and rerouting angkot to improve connectivity with BRT corridors.with the Damri corridor (Cibiru -Kebon Kalapa), so that the role of the planned corridor is no longer significant.The opportunity to reduce overlap by shifting the bus stops in the corridor is also relatively small because the position of the Damri corridor stops is also crowded.The BRT route also overlaps by 62% with the Cicaheum -Cileunyi (A10) angkot route, which serves the same route as the BRT within Bandung City, starting from Cicaheum Terminal.
b) Removing the Elang -Cicaheum BRT corridor which is less used in the quickest route between sub-districts in Bandung City, even though the start/end points are at the BRT bus stop, which is caused by the overlap with the Damri (Cibeureum -Cicaheum) and TMB (Cibeureum -Cicaheum) corridors at overlap level ~70%.
In addition, because the TMB and Damri corridors serve the same route, it is recommended to remove one of the two corridors as well.

Recommendation Scenario: Catchment area ratio improvement in the eastern area of Bandung City
This region of Mandalajati, Ujung Berung, and Cibiru sub-districts in the eastern area has a low catchment area ratio, even though it has a decent population density and low generation and attraction demand, so the integration of the public transport network does not fulfil the concept of equity.Therefore, angkot as a feeder needs to play its role in reaching users in areas that are not served by the main modes.The following is a recommended scenario for the proposed angkot rerouting.The number of transfers for trips to and from Cibeunying Kaler exceeds the maximum limit of 2 [11].
There are no bus modes that serve in the Cibeunying Kaler area, including the BRT plan.Planning that does not pay attention to the area can be caused by one of the reasons for the low demand for public transportation to and from Cibeunying Kaler.Therefore, the importance of the role of angkot as a feeder that can feed users to and from the area without having to transfer too often.
Connectivity between corridors and between modes requires that there are routes that overlap or coincide with each other, or adjacent stops, to ensure the availability of routes for trips and minimize walking.There is only one angkot route to Cibeunying Kaler that crosses the BRT corridor in the Dago area, namely the K34 Caringin -Sadang Serang route.Other angkot routes also have the potential to cross the Cibeunying Kaler area and improve angkot connectivity with BRT, namely the K33 Gedebage -Simpang Dago corridor.Therefore, the following is a proposed transfer of the K33 route to improve the connectivity of angkot with BRT.f) Rerouting Gedebage -Simpang Dago (K33) from the Surapati (south) route which does not intersect with the BRT stop to the Tubagus Ismail route which can intersect with the BRT bus stop in Dago.Thus, users to and from Cibeunying Kaler have other options and do not rely entirely on the Caringin -Sadang Serang route to continue their journey by BRT.After testing with the Network Analyst feature, it was proven that after the K33 angkot rerouting, the number of corridor changes for trips from 7 sub-districts to Cibeunying Kaler decreased from 3 times to 2 times, namely trips from the eastern area (Cibeunying Kidul, Mandalajati, Ujung Berung, Cibiru, and Penyileukan) due to the increase in angkot options to Cibeunying Kaler, and from the South Bandung area (Bojongloa Kidul and Babakan Ciparay) due to the increasing use of the Dago -Leuwipanjang BRT corridor.

Conclusions
Based on the evaluation and analysis of the public transit network, the conclusion can be drawn from this study.
• Based on spatial analysis (service area) and O-D Matrix trip analysis to assess the integration of the Greater Bandung BRT Trans Metro Pasundan network to the existing modes of Angkot, Damri Bus, Trans Metro Bandung, and KRD or commuter train in Bandung City, several conclusions can be drawn that the unclear intermodal hierarchy causes 1) high overlap level and 2) low area coverage; and low connectivity causes 3) high intermodal transfers and results in low demand with transit.These issues do not meet the concept of equity set out in the guidelines and objectives of public transport itself.• With a catchment area radius of 200 m along the angkot routes and 400 m at BRT, Damri, TMB, and KRD stops, the overlap level of the Derwati -Ciwastra -Cijerah BRT corridor towards the same angkot route (Derwati -Ciwastra -Cijerah) is 63%.Concerning other major bus modes, the overlap level of the catchment area of the BRT stops for the Alun-Alun -Cibiru corridor even reaches 92.5% over the Cibiru Damri -Kebon Kalapa corridor.When compared to the existing overlap of modes vs. angkot, BRT dominates at an overlap level of >40%, where there is 12.8% of the total combination of BRT route pairs vs. other main modes (Damri, TMB, and KRD), 3.2% for BRT vs. intercity angkot, and 1.6% for route pairs vs. angkot, compared to overlapping of existing modes vs. angkot which only reached 0.7% for angkot and 1.1% for intercity angkot.The results of this analysis show the unclear hierarchy of public transportation modes in Bandung City, especially the BRT plan towards angkot modes as feeders and Damri, TMB, and KRD Bus modes as the main modes.• Due to the overlap, BRT does not increase the coverage area in some areas that have a low catchment area ratio, such as 3 sub-districts in eastern Bandung, namely Mandalajati, Ujung Berung, and Cibiru, which only have public transport coverage of 13% -27% of the available road network, despite having a relatively low population density (~10,000 -15,000 people/km 2 ) which needs to be supported in order to thrive.Likewise, the southern part of Bandung, namely Panyileukan and Rancasari with a catchment area ratio of 44%, even though they have a relatively high demand for generation/attraction demand (~500 -700 PCU/hour), so they need to be supported so as not to trigger congestion.This shows the lack of role of angkot as feeder transport in the area.• With Network Analyst in ArcGIS, the quickest route between sub-districts is issued with a start/end point at the BRT stop to review trip indicators.After the existence of BRT to complement the existing modes of angkot, Damri, TMB, and KRD, there are still 26 out of 900 trips (~3%) that require more than 2 transfers.The 26 trips that require transfers 3 times are dominated by the destinations of Cibeunying Kaler.Other trips are from the southern part of Bandung, which includes the areas of Bojongloa Kaler and Bojongloa Kidul, as well as Buah Batu.When related to the O-D projection, the demand for transit to Cibeunying Kaler and this trip is low or zero.Twenty (20) trips with the highest demand for transit in Bandung City meet the maximum number of transfers and route availability, but only two (2) use the BRT corridor in the quickest route.• Indicators of route availability are met through quantification by walking distance to reach stops in changing corridors.Of the 900 quickest routes for 30 sub-districts in Bandung City, the distance is still below the maximum limit of 500 m [16].• On the other hand, a clear hierarchy of public transportation networks is needed in Bandung City, where the Trans Metro Pasundan BRT as the main mode should serve the main transportation corridor in Bandung City, namely the trip from west to east in three areas in Bandung City, namely the north, central, and south, as well as the trip from the north (Bandung City) to the south (Soreang).Then, the role of supporting corridors from BRT and other bus modes is also needed for the trip centre in Bandung City.Angkot as feeder transportation should be a gap filler in areas that have not been served by main corridors and supporting corridors.• In this study, sample scenario recommendations have been developed to resolve the issues identified in the integration of the Bandung Raya BRT network in Bandung City: 1) Removing the Alun-Alun -Cibiru and Elang -Cicaheum BRT corridors to reduce the high BRT overlap with angkot and Damri buses that cause regional competition and some BRT corridors are underused in the quickest route; 2) Rerouting the angkot routes from Ciwastra -Ujung Berung (A25) and Gedebage -Simpang Dago (K33) to reduce the overlap level while increasing the catchment area ratio in the eastern part of Bandung City which has a low population density, generation and attraction demands; and 3) Rerouting the angkot route Gedebage -Simpang Dago (K33) to

Figure 1 .
Figure 1.Comparison of analysis of service areas and buffer areas.Source: Aslan & Kocaman, 2018

Figure 3 .
Figure 3.The 3D view of the network dataset of the public transport network in Bandung City.

Figure 4 .
Figure 4. Example of the generation of the quickest route example from sub-districts in Bandung City to Regol.

Figure 5 .
Figure 5. Population density map in Bandung City.Visualization based on data from Bandung City in Numbers (2020).The most populous sub-districts in Bandung City are marked in red in Figure 5, namely Bojongloa Kaler in the southwest part of Bandung City, followed by Astana Anyar, just to the east.The other dense areas are Cibeunying Kidul and Batununggal which are in the middle of Bandung City.The subdistricts with the lowest density are marked in dark green, dominating the eastern and southeastern areas of Bandung, namely Gedebage, Panyileukan, and Cinambo, as well as Cidadap in northern Bandung.

Figure 6 .
Figure 6.Greater Bandung BRT Corridors Development Plan.Source: Identification based on Directorate General of Land Transportation's Webinar (2021).

Figure 7 .
Figure 7. Catchment Area of Modes in Bandung City.

Figure 8 .
Figure 8. Visualization of population density towards catchment area ratio in Bandung City.

Figure 9 .
Figure 9. Visualization of traffic generation in sub-districts towards catchment area ratio in Bandung City.

Figure 10 .
Figure 10.Visualization of attraction in subdistricts towards catchment area ratio in Bandung City.

Figure 11 .
Figure 11.Visualization of overlap level between BRT corridors vs. Angkot in Bandung City.

Figure 12 .
Figure 12.Visualization of overlap level between BRT corridors vs. Damri, TMB, and KRD in Bandung City.

Figure 14 .
Figure 14.Percentage of intermodal route pairs based on overlap level.

Figure 15 .
Figure 15.Visualization of the percentage of BRT usage in quickest routes with start/end points at BRT stops.

Figure 16 .Figure 17 .
Figure 16.Visualization of the number of transfers in quickest routes with start/end points at BRT Stops

Figure 18 .
Figure 18.Examples of low catchment area ratio issues (2) and high number of transfers for trips to and from Cibeunying Kaler (3).

Figure 19 .
Figure 19.Comparison of the BRT Trans Metro Pasundan corridor to the main corridors in the Bandung City public transport network hierarchy (black lines).
4.9.Recommendation Scenario: BRT corridor removal to reduce competition due to overlap Two BRT corridors have a very high overlap level with angkot and/or Damri and TMB buses.Therefore, recommended scenarios are prepared to reduce intermodal competition while reducing network inefficiency.

Figure 20 .
Figure 20.Recommendation to remove BRT Alun-Alun Cibiru corridor (a) that overlaps with angkot Cicaheum -Cileunyi and Damri Cibiru -Kebon Kalapa.a) Removing the BRT Alun-Alun -Cibiru corridor which has an overlap of 92% of its bus stop areawith the Damri corridor (Cibiru -Kebon Kalapa), so that the role of the planned corridor is no longer significant.The opportunity to reduce overlap by shifting the bus stops in the corridor is also relatively small because the position of the Damri corridor stops is also crowded.The BRT route also overlaps by 62% with the Cicaheum -Cileunyi (A10) angkot route, which serves the same route as the BRT within Bandung City, starting from Cicaheum Terminal.

Figure 22 .
Figure 22.Recommendation to extend angkot route A25 Ciwastra -Ujung Berung (c) and reroute angkot route K33 Gedebage -Simpang Dago (d), to improve catchment area ratio.c) Extending the Ciwastra -Ujung Berung (A25) route which initially ends at the Ujung Berung Sub-Terminal, to reach Cibiru District.The route is extended until it reaches the main mode again on collector road A.H. Nasution Street so users who are fed by the A25 route can continue their journey by bus.After the extension of the route, the catchment area ratio in Ujung Berung District increased from 23.2% to 38.5%, and in Cibiru District from 13.6% to 26.9%.

Table 2 .
Modes used in quickest routes for trips with high traffic demand.

Table 3 .
Recapitulation of BRT corridors overlap level and number of usages in quickest route.