Dataset for developing optimal headway-based bus dispatching strategy during epidemic outbreaks

Abstract

This article presents the data utilized in a study focused on identifying an optimal bus dispatching strategy in light of epidemic impacts. The study specifically examines the Xi'an Xiaozhai central business district (CBD) street network, which consists of 33 major signalized intersections and 112 bus stops associated with 12 bus routes. The dataset includes details of intersection and bus stop geospatial data, street segment and intersection design, intersection signal timing plans, bus route operational properties such as dispatching frequencies, fleet sizes, loading bay capacities, and bus-specific parameters. It also encompasses data on passenger boarding and alighting counts, as well as travelers’ origin and destination (O-D) locations, routes, and departure times during three time periods: 10:00-11:00 PM, 1:00-2:00 PM, and 7:00-8:00 PM on Monday, June 7, 2021. These times represent off-peak (10:00 PM–1:00 AM the next day), adjacent-to-peak (9:00–11:00 AM, 1:00–4:00 PM, and 8:00–10:00 PM), and peak (7:00–9:00 AM, 11:00 AM–1:00 PM, and 4:00–8:00 PM) periods, respectively. Data collection involves searching government and organizational records, utilizing Alibaba Cloud's Amap platform, conducting onsite measurements, and performing a field survey. The dataset is a valuable resource for studying the integrated operations of various urban mass transit services, including buses, bus rapid transit (BRT), and fixed guideway transit, under both normal and epidemic-affected travel conditions. Additionally, it can be used to investigate multimodal integrated urban passenger services offered by automobiles, transit, ridesharing, and active transportation modes.

Specifications Table

Subject	Transportation Management
Specific subject area	Bus transit operations
Type of data	Tables and figure •Table 1: Network configuration, geometric design, and intersection signal timing plans (“DIB_Table 1 Network and signal timing data.xlsx”) •Table 2: Bus route operational parameters (“DIB_Table 2 Bus route operations data.xlsx”) •Table 3: Bus boarding and alighting counts and passenger lists by the route and stop (“DIB_Table 3 Bus route B A counts.xlsx”) •Table 4: Node classifications according to the distance from the outbreak point (“DIB_Table 4 Node classification data.xlsx”) •Table 5: The OD matrix of travelers using the bus transit routes in 5-min intervals of the hourly off-peak, adjacent-to-peak, and peak periods (“DIB_Table 5 OD Matrix.xlsx”) •Fig. 1: The geospatial layout of the Xi'an Xiaozhai CBD street network, including details of network nodes, urban street segments, bus routes, and route-specific properties (“DIB_Figure 1 Network.zip”)
How the data were acquired	The current study utilizes the Xi'an Xiaozhai central business district (CBD) street network, which comprises 33 major signalized intersections and 112 bus stops associated with 12 bus routes. The primary data items collected for this study include geospatial data on intersections and bus stops, geometric design of urban streets and intersections, intersection signal timing plans, bus route operational properties, bus-specific parameters, passenger ridership data, as well as travelers' origin and destination (O-D) locations, routes, and departure times. To gather the data details, various methods were employed, including searching government and organizational records, utilizing Alibaba Cloud's Amap platform, conducting onsite measurements, and conducting a field survey.
Data format	Raw
Description of data collection	Data on the latitude and longitude coordinates of intersections and bus stops, as well as the geometric design of urban street segments and intersections, were obtained from Alibaba Cloud's Amap Platform using its API interface and coordinate point function. Each intersection site was assigned a two-person team of graduate students to validate the data details. The field team also verified data on signal timing plans for all 33 intersections, which were obtained from government records. Data regarding bus route-specific operational properties and bus-specific parameters were extracted from the bus operations manual. A field survey was conducted to gather data on bus operations and ridership at each bus stop during off-peak, adjacent-to-peak, and peak periods of a typical weekday. The collected data included information on bus routes, bus IDs, arrival and departure times, as well as the number of boarding and alighting passengers. With the consent of participants, the survey crews also collected additional data on the O-D locations of travelers, the routes they selected, and their departure times while waiting to board. It is important to note that no personally identifiable information was collected during this process. Furthermore, the bus ventilation rate was determined by considering factors such as wind speed and the area of the air outlet within the bus's interior air conditioning system. In addition, geospatial data on the locations of intersections, bus stops, and a major hospital in the study area were utilized to identify the point of the epidemic outbreak.
Data source location	Institution: Chang'an University City/Town/Region: Xi'an Country: China
Data accessibility	Repository name: Dataset for developing optimal headway-based bus dispatching strategy during epidemic outbreaks DOI:10.17632/zc7cjr6532.2 Direct URL to data: http://dx.doi.org/10.17632/zc7cjr6532.2
Related research article	Huang, Y., Li, Z., Zhang, S., Zhou, B., Zhang, L. Optimal Headway-Based Bus Dispatching Strategy under the Influence of Epidemic Outbreaks. Sustain. Cities Soc., 92 (2023) 104468. https://doi.org/10.1016/j.scs.2023.104468.

Value of the Data

• •This dataset provides comprehensive information on intersection geospatial data, street segment and intersection geometric design, and intersection signal timing plans within a densely populated urban street network. It also includes details regarding bus transit services within the network, such as bus stop geospatial coordinates, route-specific bus dispatching frequencies, fleet sizes, loading bay capacities, bus-specific parameters, passenger boarding and alighting counts, and travelers’ O-D trip matrices.
• •Researchers and transit operators can leverage this extensive dataset to develop effective strategies for passenger services within a densely populated urban street network, considering both normal and epidemic-affected travel conditions. Additionally, this dataset can be utilized in graduate-level teaching, particularly in topics related to econometric or multivariate analysis of travel choices.
• •When combined with additional datasets containing similar attributes for other transit modes (e.g., bus rapid transit, fixed guideway transit, demand-responsive transit) as well as automobile, ridesharing, and active transportation modes, researchers, and practitioners can employ this data to develop efficient strategies for multimodal integrated passenger services in both normal and epidemic-affected travel conditions.

1. Objective

The dataset was utilized in the proposed model for computational experiments outlined in the aforementioned research article. The study findings have yielded valuable insights into the effects of various combinations of infection rates, social distancing rules, and vaccination rates on the travel time cost of all bus passengers and the cost of medical treatments for newly infected bus riders. These insights can assist bus transit operators in developing an optimal headway-based bus dispatching strategy that minimizes the weighted total cost of travel time and medical treatments.

Currently, researchers and practitioners can utilize the dataset to apply existing models to develop optimal operational strategies for urban passenger services in both normal and epidemic-affected travel conditions. Looking ahead, the dataset holds potential for experimentation and validation of new models that facilitate effective urban bus transit operations during epidemic outbreaks, considering various aspects. Firstly, the optimization of headway-based bus dispatching strategies can be extended to a rolling horizon approach, enabling the derivation of effective infection mitigation measures that account for temporal and spatial variabilities during severe epidemics. Secondly, the inclusion of demand-responsive bus services alongside regular bus services can help reduce the frequency and duration of contacts for vulnerable bus transit travelers, such as senior citizens, patients, and individuals with movement restrictions. By exploring these directions, the dataset can continue to contribute to the advancement of research and practice in developing effective urban bus transit operations under the influence of epidemic outbreaks.

2. Data Description

The dataset is organized into several tables and a zipped ArcGIS file, each containing specific information related to the Xi'an Xiaozhai CBD street network and bus transit services. Here are brief descriptions of each component:

Table 1 provides an overview of the Xi'an Xiaozhai CBD street network, covering 33 major signalized intersections and 112 bus stops along 12 bus routes. It includes intersection geospatial coordinates, configurations, and signal timing plans. Geometric design details for each intersection approach, such as lane numbers, widths, and designations for different transportation modes, are also included. The table can be accessed at http://dx.doi.org/10.17632/zc7cjr6532.2 (file name: DIB_Table 1 Network and signal timing data.xlsx).

Table 2 contains specific details of the bus transit services for the 12 bus routes. It includes information on the bus stop location, docking bay capacity, fleet size, headway, and bus volume, ventilation rate, and propulsion technology. The table is available at http://dx.doi.org/10.17632/zc7cjr6532.2 (file name: DIB_Table 2 Bus route operations data.xlsx).

Table 3 focuses on passenger boarding and alighting counts, as well as individual travelers' O-D locations, routes, and departure times. The data covers one-hour durations for off-peak, adjacent-to-peak, and peak periods. Notably, passengers boarding and alighting at bus stops near node 12 are primarily patients and their companions visiting a nearby hospital, while those at bus stops neighboring nodes 18, 19, 20, 23, and 24 are engaged in business, commercial, shopping, recreational trips, and local residents' travel. The table can be found at http://dx.doi.org/10.17632/zc7cjr6532.2 (file name: DIB_Table 3 Bus route B A counts.xlsx).

Table 4 provides distances between each node and node 12, which is identified as the outbreak point of the epidemic. Nodes are classified into three categories based on their distances to node 12 (near, medium, and far), using threshold values of 1 km and 2 km. The table is accessible at http://dx.doi.org/10.17632/zc7cjr6532.2 (file name: DIB_Table 4 Node classification data.xlsx).

Table 5 contains O-D matrices, representing passenger travel from origin to destination zones within one-hour durations of off-peak, adjacent-to-peak, and peak periods. The trips are aggregated for approximately 5-min intervals. Each cell in the table represents the number of passengers traveling between specific origin and destination zones. The table can be downloaded from http://dx.doi.org/10.17632/zc7cjr6532.2 (file name: DIB_Table 5 OD Matrix.xlsx).

Fig. 1 is a visual representation of the Xi'an Xiaozhai CBD network, illustrating the layout of network nodes (intersections), street segments, and bus routes. It also highlights the 1-km and 2-km boundaries from node 12, the identified epidemic outbreak point. The file is provided in a compressed format (DIB_Figure 1 Network.zip) and includes data details of node coordinates, geometric design, and bus operations, compatible with ArcGIS maps (in formats such as MDX, shp, and dbf). It can be accessed at http://dx.doi.org/10.17632/zc7cjr6532.2.

Fig. 1.

Illustration of the epidemic outbreak point, nodes adjacent to high travel intensity bus stops, and ranges of near-, medium-, and far-distance nodes.

3. Experimental Design, Materials, and Methods

3.1. Study area network nodes neighboring high travel intensity bus stops

A field survey was conducted by survey crews comprising graduate and undergraduate students led by faculty members and researchers. One week prior to the field survey, all students received basic training regarding the purpose of the survey, the full autonomy and privacy of travelers' participation, an electronic survey form with survey questions, and information on data collection, processing, and confidentiality. On the day of the field survey, two students were assigned to each bus stop during specific periods: 10:00–11:00 AM, 1:00–2:00 PM, and 7:00–8:00 PM for data collection. These periods represented off-peak (10:00 PM–1:00 AM the next day), adjacent-to-peak (9:00–11:00 AM, 1:00–4:00 PM, and 8:00–10:00 PM), and peak (7:00–9:00 AM, 11:00 AM–1:00 PM, and 4:00–8:00 PM) periods, respectively.

During the data collection, the survey crews mounted an HD camera on a tripod stand at each bus stop to record bus routes, bus IDs, arrival and departure times, and boarding and alighting passengers. With prior consent, the survey crews also collected additional data on travelers’ O-D locations, selected routes, and departure times of travelers waiting to board. The interactive process of the travel choice data collection ensured full autonomy and privacy of travelers' participation without collecting any personally identifiable information. All raw data extracted from the video clips and electronic survey forms were initially stored in MS Excel files on password-protected laptops. These files were then compiled and transferred to an offline USB unit, which was securely stored in a locked drawer to maintain confidentiality. The survey data were used to create bus route-specific, stop-based boarding and alighting counts, and O-D trip matrices corresponding to one-hour durations of off-peak, adjacent-to-peak, and peak periods.

Analysis of bus passenger travel data revealed that bus stops near nodes 12, 18, 19, 20, 23, and 24 experienced high travel intensities. To provide further insights, Table 6 displays the ridership counts specific to bus routes at the neighboring bus stops.

Table 6.

The study area network nodes adjacent to high travel intensity bus stops.

Period			Alighting count (persons)				Boarding count (persons)
Node	Bus route	Bus stop	Off-peak	Adjacent-to-peak	Peak	Daily	Off-peak	Adjacent-to-peak	Peak	Daily
12	7	('19′, '12′)	220	346	436	6570	19	42	50	751
	7	('3′, '12′)	36	57	72	1083	180	253	278	4535
	9	('11′, '12′)	118	165	199	3101	127	202	215	3515
	9	('13′, '12′)	143	218	291	4283	93	133	171	2578
18	1	('17′, '18′)	20	28	36	544	206	302	391	5860
	1	('19′, '18′)	254	384	496	7418	20	28	36	544
	2	('16′, '18′)	138	269	324	4889	100	146	178	2746
	2	('27′, '18′)	129	176	233	3483	117	156	172	2819
	6	('17′, '18′)	24	32	36	584	78	132	161	2446
	6	('19′, '18′)	108	136	180	2716	60	86	108	1646
19	1	('18′, '19′)	99	155	218	3126	230	326	302	5388
	1	('20′, '19′)	231	322	416	6275	97	140	173	2655
	6	('18′, '19′)	76	108	122	1960	75	111	135	2082
	6	('20′, '19′)	128	201	247	3767	58	88	108	1654
	7	('12′, '19′)	180	291	342	5313	201	322	370	5817
	7	('23′, '19′)	286	471	568	8699	167	248	292	4573
20	1	('19′, '20′)	285	439	606	8776	20	28	36	544
	1	('21′, '20′)	57	77	106	1558	165	239	262	4264
	3	('13′, '20′)	98	182	204	3200	115	165	191	3028
	3	('24′, '20′)	206	243	321	4887	123	159	207	3138
	11	('13′, '20′)	59	96	135	1929	23	36	38	625
	11	('24′, '20′)	11	15	19	290	141	203	222	3620
23	7	('19′, '23′)	220	348	423	6480	52	82	114	1642
	7	('28′, '23′)	80	101	127	1963	178	263	250	4375
24	3	('20′, '24′)	200	316	424	6204	30	45	54	837
	3	('32′, '24′)	11	15	21	306	185	261	288	4686
	11	('20′, '24′)	24	38	44	690	38	63	85	1235
	11	('32′, '24′)	65	103	140	2036	20	30	36	558

Note: Daily boarding or alighting counts were estimated as the weighted total of 3-hour off-peak, 7-hour adjacent-to-peak, and 8-hour peak period boarding or alighting counts.

3.2. Classification of study area network nodes by distance from the outbreak point

Among the nodes adjacent to high-travel-intensity bus stops, node 12 stands out due to its proximity to bus stops where a considerable number of passengers, primarily patients and their companions visiting a nearby hospital, board and alight. Consequently, node 12 is identified as a high-risk area for epidemic outbreaks. In the study area, the remaining nodes are classified as near, medium, or far distance nodes from the outbreak point (node 12) based on 1-km and 2-km radii. The study assumes, as mentioned in [1], that the risk of infection at nodes located near, medium, and far distances from the outbreak point decreases with increasing distance. Table 7 provides the node classifications, assigning descending levels of infection risks based on their proximity to the outbreak point. The source file for Table 7, "DIB_Table 4 Node classification data.xlsx," can be accessed at http://dx.doi.org/10.17632/zc7cjr6532.2. Additionally, Fig. 1 illustrates the node locations, street network connections, the outbreak point, and the ranges of near-, medium-, and far-distance nodes in the study area. The compressed file folder for Fig. 1, ``DIB_Figure 1 Network.zip,'' is available at http://dx.doi.org/10.17632/zc7cjr6532.2.

Table 7.

Classifications of near-, medium-, and far-distance nodes in the study area network based on the distance to node 12 as the epidemic outbreak point.

Node	Classification	Node	Classification
1	Medium	17	Medium
2	Near	18	Near
3	Near	19	Near
4	Medium	20	Near
5	Medium	21	Medium
6	Medium	22	Medium
7	Near	23	Near
8	Near	24	Medium
9	Medium	25	Medium
10	Medium	26	Medium
11	Near	27	Medium
12	Near	28	Near
13	Near	29	Medium
14	Medium	30	Medium
15	Medium	31	Medium
16	Near	32	Medium
		33	Far

Ethics Statements

The authors declare that the work conducted in this research did not involve the use of animal experiments or social media data. Furthermore, all authors have carefully reviewed and agreed upon the last version of this manuscript.

CRediT authorship contribution statement

Yan Huang: Conceptualization, Methodology, Data curation, Software, Writing – original draft. Zongzhi Li: Methodology, Validation. Shengrui Zhang: Investigation. Bei Zhou: Data curation, Validation, Writing – review & editing. Lei Zhang: Software.

Data Availability

• Dataset for developing optimal headway-based bus dispatching strategy during epidemic outbreaks (Original data) (Mendeley Data).