Bus transportation holds the primary position of public transportation in Indian cities after introducing Jawaharlal Nehru National Urban and Rural Mission scheme. The significant rise in passenger demand over recent decades has increased the need for improved vehicular and passenger facilities at bus stations. This study analysed the factors affecting the performance of Arapalayam Bus Station in Madurai, which serves both intercity and local bus services. The field investigations revealed that current service intensity is insufficient to accommodate the bus arrival rate, leading to queuing outside the station premises. This results in traffic congestion on city roads, develop conflicts between vehicles-pedestrians and increased delay in bus services. To evaluate the impact of each factor, a questionnaire survey was conducted and responses were ranked using a statistical method known as principal component analysis. The results indicated that absence of bay optimisation, separate bay facilities, streamlined passenger movement and bus queuing significantly affects station operations. In addition, absence of dedicated feeder service leads to passenger pick-up and drop-off near the adjacent carriageways, further contributing to blockages and congestion. The study designed an integrated bus terminal layout with improved feeder facilities to achieve multi-modal traffic operations in practice.
Notation
1. Introduction
A bus station is a major infrastructure facility designed to accommodate a large number of buses and passengers in response to growing demand (Kosmidis and Müller-Eie, 2024; Siddiqui et al., 2024; Logeswaran et al., 2022). A recent investigation at the Arapalayam Bus Station in Madurai revealed that the absence of standard operating procedures significantly reduces the station’s operational capacity (Sharma et al., 2019). The key factors contributing to this loss include inadequate transit holding areas (Siddiqui et al., 2024), poor pedestrian infrastructure (Vigneshwaran et al., 2022; Sarkar et al., 2017; Jain, 2015) and lack of feeder services. These shortcomings lead to service delays, traffic blockages and reduced service intensity, ultimately resulting in vehicle queuing and overall inefficiency in station operations. This bus station is located in the heart of Madurai city and serves as a gateway to southern Tamil Nadu. As it operates around the clock with an average of 80–100 bus movements per hour, there is a critical need for upgraded infrastructure and enhanced passenger amenities to meet the growing demand efficiently (Comert et al., 2024). Therefore, the researchers have decided to conduct a questionnaire and traffic survey to identify the factors affecting bus station traffic operations using both manual and mechanical methods. The traffic study was planned during both the lean and summer seasons to assess the consistency of influencing factors across different periods (Akbulut et al., 2022).
A dedicated team was formed to carry out direct field investigations by visiting the site to observe operational issues. Based on these observations, a structured questionnaire was developed with experts feedback to evaluate the intensity of the identified factors through stakeholder’s feedback (Ferencz and Zöldy, 2023; Mahmoudi et al., 2024). A comprehensive review of existing literature was conducted to identify the key factors influencing bus station performance across various contexts, along with the tools and methodologies employed to address the prevailing challenges. The effective planning and design of bus terminals play a vital role in enhancing operational efficiency, passenger convenience and the overall sustainability of urban transportation systems. Many recent studies have examined these aspects from various perspectives. Handayani et al. (2024) and Buulolo et al. (2023) highlighted the significant influence of government policies and user convenience on the effectiveness of bus stations, even in conditions of limited accessibility. Ahmad et al. (2024) investigated the spatial configuration of BRT stations by integrating accessibility, platform design and passenger flow dynamics with urban land use. Their findings showed that strategic station placement could reduce travel time by 15% and improve platform efficiency by 25%. In a regional context, Birhan (2024) applied queuing theory to public transport stations in South West Ethiopia, emphasising how fluctuations in demand, layout design and vehicle scheduling impact service quality. Boadi-Kusi et al. (2024) focused on accessibility challenges for visually impaired individuals, identifying infrastructure limitations and advocating for inclusive design and supportive transport policies. The issue of shared urban transport spaces was addressed by Zhang et al. (2022) who found that certain bus stop designs increase conflict between passengers and cyclists, recommending spatial separation and traffic-calming measures.
Ramos-Santiago (2022) established a strong correlation between walkability around feeder bus stops and increased boardings at rapid transit stations, highlighting the importance of pedestrian-oriented planning. Luo et al. (2022) developed a model to estimate bus queuing delays caused by blocking effects at curbside stops, offering optimisation strategies to improve operational flow. Li (2021) presented a case study on terminal layout redesign incorporating sustainability and space optimisation. In addition, Kadiyali (2024) provided a comprehensive framework for bus terminal planning, including site selection, layout efficiency and environmental considerations. Collectively, these studies emphasise a data-driven and user-centric approach to design efficient and inclusive bus terminals.
1.1 Research gap
While several studies have identified factors related to traffic operations and government policies, the operational requirements of bus stations in response to varying traffic demand, multi-modal integration and land use pattern remain inadequately addressed. This gap presents a significant challenge to achieve functional efficiency. Accordingly, the research gaps have been identified based on this critical aspect.
Land use efficiency of the existing station
Assessment of traffic handling capacity
Service intensity and queuing rate of waiting buses
Streamlined layout to handle the growing traffic demand with defined traffic movements
Congestion analysis for the existing station to observe potential conflict points
1.2 Objectives of the study
Based on the research gaps identified, the objective of the present study is to:
Find the key factors influencing the performance and functionality of bus stations.
Rank the factors affecting the design and efficiency of existing bus station layouts.
Develop an optimised station layout with multi-modal transport facility.
2. Methodology
The study commenced with a comprehensive field investigations aimed to identify the key factors influencing traffic operations at the study area. These factors were used to develop a questionnaire. Furthermore, a questionnaire survey was conducted along with traffic survey focusing on important metrics such as traffic volume, congestion, service intensity, peak-hour traffic, queuing, capacity, parking, traffic demand, pedestrian movement (Quijada-Alarcón et al., 2025), conflict, feeder service and amenities. The response of the factors is then converted into normalised matrix to assess the intensity of the factors. The ranking of factors is then observed using principal component analysis (PCA) analysis. Furthermore, queuing analysis was done to observe to check the service intensity of the station. Then, by addressing all these factors with spatial design for each component, a new layout is proposed with all the passenger amenities with passenger amenities according to the growing demand. Figure 1 shows the flowchart of the research methodology.
The process begins with the start node leading to field
investigation, where passengers, vehicles, and service intensity are
observed. It proceeds to factor identification followed by a questionnaire
survey. The data collected are standardised into a normalised matrix, then
analysed through principal component analysis. The analysis results include
queuing and congestion assessments, which inform layout design and
subsequent spatial design, concluding with the end node. The flow shows an
iterative, data-driven approach to developing a functional terminal
layout.Flowchart of the proposed methodology
The process begins with the start node leading to field
investigation, where passengers, vehicles, and service intensity are
observed. It proceeds to factor identification followed by a questionnaire
survey. The data collected are standardised into a normalised matrix, then
analysed through principal component analysis. The analysis results include
queuing and congestion assessments, which inform layout design and
subsequent spatial design, concluding with the end node. The flow shows an
iterative, data-driven approach to developing a functional terminal
layout.Flowchart of the proposed methodology
3. Study area
This section specified the geographical specifications of the study area. Table 1 gives the detailed specifications include station name, geographical coordinators, area, perimeter of the terminal boundary and the type of service provided whereas Figure 2 provides the layout boundary using Google earth measurements.
Specifications of the study area
| Station name | Geographical indication | Area: m2 | Perimeter: m | Type of service |
|---|---|---|---|---|
| Madurai Arapalayam | 9.9347°N 78.1036°E | 7237 | 361 | Mofussil and city |
The image displays a satellite view of Madural Arapalayam with the
Arapalayam Bus Stand marked in blue. The bus stand is centrally located,
depicted within the outlined area, surrounded by various buildings and
greenery in the surrounding area. The overall layout includes a mix of
residential and commercial structures. Key roads are visible nearby, and the
scene reflects an urban environment with a clear depiction of the bus
stand's placement relative to other structures. Spatial relationships among
buildings and the bus stand are evident, allowing for an understanding of
the area's layout.Boundary of Madurai Arapalayam bus station
The image displays a satellite view of Madural Arapalayam with the
Arapalayam Bus Stand marked in blue. The bus stand is centrally located,
depicted within the outlined area, surrounded by various buildings and
greenery in the surrounding area. The overall layout includes a mix of
residential and commercial structures. Key roads are visible nearby, and the
scene reflects an urban environment with a clear depiction of the bus
stand's placement relative to other structures. Spatial relationships among
buildings and the bus stand are evident, allowing for an understanding of
the area's layout.Boundary of Madurai Arapalayam bus station
4. Factors influencing station’s performance
Through direct field investigation, the study observed the factors (listed in Table 2) affecting the traffic operations (Logeswaran et al., 2025). The factors are classified based on major five metrics such as capacity, space requirements and functional attributes, space restrictions for bus bay operations, passenger traffic and absence of feeder service.
Factors affecting station’s performance
| Code | Metrics | Factor | Observations |
|---|---|---|---|
| 1 | Traffic capacity | Land use | The formulation of a well-structured land use is essential to facilitate seamless accessibility for pedestrians, buses and other vehicles approaching the station. The roads adjacent to the bus station experience significant congestion, particularly at entry and exit points, leading to additional delays and disruptions in regular traffic flow. |
| 2 | — | Average layover time, insufficient bus circulation area and Queuing | Bus queuing is primarily caused due to random arrival rates and relatively constant departure rates, a condition further worsened by extended layover durations and insufficient bay allocation for specific routes. The minimum clearance between two buses is significantly reduced, contributing to congestion and operational inefficiencies. Moreover, during peak hours, the sharp increase in arrival rates for particular routes results in prolonged queuing and extended passenger waiting times. |
| 3 | — | Demand analysis / Service frequency / Traffic forecasting / Bay inadequacy | The frequency of both city and long-route bus services should be accurately identified and forecasted during the planning stage to ensure the provision of an adequate number of bays. Inadequate bay allocation results in extended service times and operational delays. |
| 4 | Space requirements and location of functional attributes | Area distribution and Passengers movement | Absence of specific area distribution such as site area, open area and at-grade parking leads to poor land use and passenger movement regulation is not in a regulated way, i.e. poor pedestrian traffic management |
| 5 | — | Poor traffic management / Station entry or exit | The lack of essential traffic regulations such as designated pedestrian pathways and separate lanes for buses and other vehicles contribute to ineffective traffic management within and around bus terminals. Furthermore, poorly designed terminal entry and exit points result in congestion and blockages on adjacent roads, further disrupting the overall traffic flow. |
| 6 | — | Conflicts (passengers-vehicles) | Passenger movement within the bus circulation area often leads to conflicts between pedestrians and moving buses. |
| 7 | Space restrictions for bus bay operations | Absence of ITS | Intelligent transportation system technologies such as passenger guidance systems, vehicle guidance systems, traffic management systems, real-time bus information displays and standardised operating procedures for bus stations are largely absent in the current context. |
| 8 | — | Optimisation of bus bays | To prevent bus bunching and streamline loading, unloading and shunting operations, advanced information systems such as SMS alerts for drivers and real-time display boards should be implemented for better coordination and operational efficiency. |
| 9 | — | Absence of separate bay operations for passenger loading /unloading and bus shunting (ideal) activities | Unlike USA and UK, Indian bus stations often lack designated areas for passenger unloading, bus loading, and shunting operations. Instead, these activities are typically carried out in a common space, leading to frequent reverse movements of buses, operational blockages and an increased risk of minor accidents, often attributed to limited manoeuvring space and driver negligence. |
| 10 | Passenger traffic | Access constraints of passenger amenities (Sedaya and Sulandari, 2019; Doddi et al., 2024) | Passenger facilities should be systematically organised to ensure a continuous and regulated flow of movement within the terminal. The improper placement or dislocation of these facilities often leads to uncoordinated and uncontrolled passenger movement. To enhance accessibility and operational efficiency, such facilities must be strategically located in close proximity to the bus layover locations. |
| 11 | — | Platform and access area, passenger handling capacity (Angadi et al., 2024) | The spatial design of platforms involves key considerations such as the arrangement of bus bays, capacity to handle peak passenger volume, provision for bus layover and dedicated facilities for drivers. |
| 12 | Absence of feeder service facilities | Kiss-n-Ride sharing, city bus station, taxi stand, auto stand and parking | The absence of dedicated feeder service often results in unorganised passenger drop-off and pick-up within the terminal area, leading to delays in bus operations. Furthermore, several terminals lack designated parking areas for passenger vehicles due to space constraints. In some cases, both terminal operators and passengers are forced to share the same limited parking space, further contributing to congestion and operational inefficiencies. |
5. Data collection
5.1 Questionnaire form
Table 3 presents the questionnaire form used for feedback collection from the stakeholders. It comprises of 25 questions designed to cover all relevant factors as listed in Table 2. The final version of questionnaire was developed based on expert feedback prior to distribution. A total of 100 responses were received from a diverse stakeholders including drivers, conductors, station operators, commercial shop owners and local residents. The respondents were asked to rate each question using a Likert scale ranging from 1 to 5, where 1 indicates ‘Excellent’, 2 ‘Good’, 3 ‘Average’, 4 ‘Poor’ and 5 ‘Very Poor’.
Questionnaire form
| Questions | Factor | Sub factor |
|---|---|---|
| Does the existing land use plan able to serve for large volume of vehicular and pedestrian traffic? | 1 | Land use plan |
| Do you think that the location of the bus station is at the correct place of a city to reach easily? | 1 | Location |
| Do you think surrounding land use activity affects the bus station operations? | 1 | Surrounding land use |
| Whether the average layover time of individual routes sufficient to handle large number of buses? | 2 | Increasing average layover time |
| Does this station has sufficient bus circulation area? | 2 | Insufficient bus circulation area |
| What is the intensity of queuing buses? | 2 | Queuing |
| Whether the bus station is designed to handle future traffic? | 3 | Traffic forecasting |
| Does service intensity affects the layover time? | 3 | Service frequency |
| Whether the bus station is designed with sufficient bay allocation of routes? | 3 | Inadequacy of bay |
| Whether the area distribution is sufficiently provided to handle vehicle and passenger traffic volume? | 4 | Area distribution and irregular passenger movement |
| Does the bus station follows proper traffic management? | 5 | Improper traffic management |
| Whether the bus station has sufficient entry width and exit width for buses from the adjacent roads? | 5 | Station entry or exit |
| Rate the intensity of passenger–vehicle conflicts | 6 | Passenger–vehicle conflicts |
| ITS technology | 7 | Absence of ITS technologies |
| Optimisation of bus bay operations | 8 | Optimisation of bays |
| Whether separate bay for loading, unloading and shunting of buses are available? | 9 | No separate bay facilities |
| Does the station has access constraint in passenger facilities? | 10 | Access constraints of passenger amenities |
| Whether platform designed is sufficient to handle large volume of passengers? | 11 | Platform and access area |
| Rate the passenger handling capacity | 11 | Passenger handling capacity |
| Service facility of ‘Kiss-n-Ride’ parking | 12 | Kiss-n-Ride parking, city bus station, taxi station, auto station, private vehicles congestion and parking |
| Service facility of city bus station | 12 | — |
| Whether sufficient space provided for taxi station? | 12 | — |
| Whether sufficient space provided for auto stand? | 12 | — |
| Congestion due to private vehicles | 12 | — |
| Whether sufficient parking facility is available for two wheeler and cars for both the station staffs and passengers? | 12 | — |
| Any other points/Suggestions for improvement: | ||
The data received through responses are converted into normalised matrix form representing the factors on various years. Table 4 represents the normalised matrix sheet, where row represent the factors (12) and column represents the years (5).
5.2 Traffic survey
A reliable and comprehensive data collection methodology is essential for accurately evaluating layout operations (Zuo et al., 2025; Wang et al., 2020). Accordingly, a traffic study was carried out over a five-year period from 2019 to 2024. To capture seasonal variations in traffic flow, surveys were conducted twice annually, once during the harvest season and once during the lean season. Based on these observations, the average maximum hourly traffic volume was recorded for each day, from which the peak hour traffic was identified (Wei et al., 2024; Tsuboi, 2021). Table 5 represents the consolidated traffic data during the observation period.
Traffic data (collective response)
| Time: min | Arrival rate | Departure rate | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| R1 | R2 | R3 | R4 | R5 | R6 | R7 | R8 | R9 | R1 | R2 | R3 | R4 | R5 | R6 | R7 | R8 | R9 | |
| 12.00–13.00 | 12 | 8 | 6 | 12 | 9 | 11 | 10 | 16 | 18 | 6 | 3 | 6 | 6 | 4 | 8 | 4 | 12 | 14 |
| 13.00–14.00 | 9 | 11 | 8 | 15 | 8 | 12 | 6 | 15 | 16 | 5 | 2 | 6 | 8 | 4 | 8 | 4 | 12 | 12 |
| 14.00–15.00 | 11 | 12 | 9 | 13 | 9 | 14 | 11 | 16 | 15 | 4 | 5 | 6 | 6 | 4 | 6 | 4 | 12 | 10 |
| 15.00–16.00 | 9 | 10 | 6 | 16 | 10 | 14 | 9 | 18 | 15 | 4 | 3 | 6 | 10 | 4 | 10 | 4 | 12 | 12 |
| 16.00–17.00 | 10 | 6 | 9 | 12 | 8 | 17 | 10 | 16 | 15 | 5 | 3 | 6 | 6 | 4 | 6 | 4 | 12 | 9 |
| 17.00–18.00 | 11 | 9 | 12 | 10 | 9 | 10 | 8 | 15 | 16 | 4 | 3 | 6 | 6 | 4 | 6 | 5 | 12 | 12 |
| 18.00–19.00 | 8 | 12 | 9 | 12 | 10 | 12 | 11 | 16 | 18 | 6 | 3 | 6 | 6 | 4 | 6 | 4 | 12 | 10 |
| 19.00–20.00 | 8 | 9 | 8 | 9 | 9 | 12 | 9 | 12 | 16 | 4 | 3 | 6 | 10 | 4 | 8 | 4 | 12 | 12 |
| 20.00–21.00 | 6 | 9 | 8 | 11 | 9 | 10 | 4 | 10 | 15 | 5 | 3 | 6 | 6 | 4 | 8 | 3 | 12 | 14 |
| 21.00–22.00 | 9 | 5 | 9 | 14 | 9 | 14 | 2 | 16 | 15 | 4 | 3 | 5 | 8 | 4 | 8 | 6 | 12 | 12 |
| 22.00–23.00 | 6 | 5 | 4 | 12 | 4 | 12 | 4 | 15 | 16 | 5 | 4 | 6 | 9 | 4 | 9 | 4 | 12 | 9 |
| 23.00-0.00 | 6 | 4 | 2 | 9 | 5 | 11 | 5 | 16 | 15 | 4 | 3 | 4 | 6 | 4 | 6 | 4 | 12 | 12 |
| 0.00–01.00 | 2 | 5 | 5 | 6 | 6 | 6 | 6 | 18 | 14 | 2 | 3 | 2 | 9 | 4 | 9 | 4 | 12 | 15 |
| 01.00–02.00 | 3 | 4 | 1 | 9 | 5 | 10 | 8 | 17 | 12 | 2 | 3 | 6 | 9 | 4 | 9 | 5 | 12 | 12 |
| 02.00–03.00 | 2 | 3 | 1 | 12 | 4 | 12 | 4 | 16 | 10 | 1 | 3 | 6 | 9 | 4 | 10 | 4 | 12 | 9 |
| 03.00–04.00 | 6 | 3 | 1 | 10 | 3 | 10 | 3 | 15 | 15 | 5 | 4 | 6 | 6 | 4 | 8 | 4 | 12 | 12 |
| 04.00–05.00 | 5 | 6 | 5 | 12 | 6 | 9 | 6 | 14 | 15 | 4 | 3 | 6 | 6 | 4 | 6 | 4 | 12 | 12 |
| 05.00–06.00 | 9 | 9 | 4 | 9 | 9 | 9 | 8 | 13 | 15 | 3 | 3 | 6 | 8 | 4 | 8 | 4 | 12 | 15 |
| 06.00–07.00 | 9 | 9 | 6 | 6 | 9 | 6 | 9 | 16 | 14 | 4 | 3 | 5 | 7 | 4 | 7 | 4 | 12 | 12 |
| 07.00–08.00 | 10 | 10 | 9 | 12 | 12 | 12 | 0 | 15 | 16 | 4 | 3 | 6 | 6 | 4 | 6 | 4 | 12 | 10 |
| 08.00–09.00 | 9 | 8 | 6 | 14 | 9 | 12 | 9 | 16 | 15 | 4 | 3 | 6 | 7 | 4 | 7 | 4 | 12 | 12 |
| 09.00–10.00 | 9 | 15 | 6 | 12 | 10 | 12 | 10 | 17 | 12 | 4 | 3 | 6 | 6 | 4 | 6 | 4 | 12 | 12 |
| 10.00–11.00 | 8 | 6 | 9 | 12 | 11 | 15 | 8 | 15 | 15 | 5 | 3 | 6 | 7 | 4 | 7 | 4 | 12 | 11 |
| 11.00–12.00 | 9 | 9 | 9 | 12 | 12 | 12 | 12 | 16 | 16 | 6 | 4 | 6 | 7 | 4 | 7 | 4 | 12 | 10 |
R1–R9 represent service routes. The details of the routes are R1 – Periyakulam, R2 – Theni, R3 – Cumbum, R4 – Erode, R5 – Dindigul, R6 – Salem, R7 – Palani, R8 – Coimbatore, R9 – Tiruppur
Figure 3 illustrates the hourly variation of bus arrival and departure rates. The maximum number of buses utilised the station in a hour is considered as peak hour. To determine the peak value, bus arrival rates were recorded on an hourly basis and further segmented into 10-minute intervals (e.g. 8:00 a.m. to 9:00 a.m. was divided into 8:00–8:10 a.m., 8:10–8:20 a.m., etc.). On the observed maximum day, the peak hour for arrivals occurred between 2:00 and 3:00 p.m., while the peak departure hour was between 3:00 and 4:00 p.m. The overall peak hour with the highest average activity across the observation period was identified as 3:00–4:00 p.m. The terminus has a capacity to accommodate a maximum of six buses in a row at any given time, including those in the boarding phase. Tables 6 and 7 represent the maximum service intensity and arrival/departure rate of peak hour respectively.
This bar graph illustrates the number of arrivals and departures
within various hourly intervals, ranging from twelve to one o'clock in
the morning. The horizontal axis lists time intervals in hours, while
the vertical axis measures the counts from zero to one hundred and
twenty. Solid black bars represent arrivals, and patterns marked with
diagonal lines indicate departures. The data flows left to right, with
hour intervals ordered sequentially. Each pair of bars for every hour
visually compares the number of arrivals and departures, facilitating an
understanding of the overall traffic during these times. The highest
points for arrivals and departures vary throughout the presented hours,
creating a pattern that reflects hourly activity levels.Hourly arrival and departure rate of buses
This bar graph illustrates the number of arrivals and departures
within various hourly intervals, ranging from twelve to one o'clock in
the morning. The horizontal axis lists time intervals in hours, while
the vertical axis measures the counts from zero to one hundred and
twenty. Solid black bars represent arrivals, and patterns marked with
diagonal lines indicate departures. The data flows left to right, with
hour intervals ordered sequentially. Each pair of bars for every hour
visually compares the number of arrivals and departures, facilitating an
understanding of the overall traffic during these times. The highest
points for arrivals and departures vary throughout the presented hours,
creating a pattern that reflects hourly activity levels.Hourly arrival and departure rate of buses
Maximum service intensity/day
| R | R1 | R2 | R3 | R4 | R5 | R6 | R7 | R8 | R9 |
|---|---|---|---|---|---|---|---|---|---|
| AR | 11 | 12 | 9 | 13 | 9 | 14 | 11 | 16 | 15 |
| DR | 7 | 6 | 6 | 8 | 5 | 6 | 5 | 12 | 10 |
| MQC | 6 | 6 | 6 | 6 | 6 | 6 | 6 | 6 | 6 |
| Time = 2.00 p.m.–3.00 p.m., No. of arrivals = 110, No. of departures = 65 | |||||||||
Arrival and departure rate/h
| R | R1 | R2 | R3 | R4 | R5 | R6 | R7 | R8 | R9 |
|---|---|---|---|---|---|---|---|---|---|
| AR | 12 | 15 | 12 | 16 | 12 | 17 | 12 | 18 | 18 |
| DR | 7 | 6 | 6 | 10 | 5 | 10 | 6 | 12 | 15 |
| Difference | 5 | 9 | 6 | 6 | 7 | 7 | 6 | 6 | 3 |
| MQC | 6 | 6 | 6 | 6 | 6 | 6 | 6 | 6 | 6 |
From Table 7, the capacity of the queue exceeds for R2, R5 and R6. Therefore, one-third of the total routes exceeded its capacity. Table 8 lists the layover time of each route, whereas Figure 4 illustrates the bar graph of LT.
Layover time
| R | WT: min | OT: min | LT: min |
|---|---|---|---|
| Periyakulam | 10 | 50 | 60 |
| Theni | 20 | 100 | 120 |
| Cumbum | 10 | 50 | 60 |
| Erode | 9 | 45 | 54 |
| Dindigul | 15 | 75 | 90 |
| Salem | 8 | 40 | 48 |
| Palani | 15 | 75 | 90 |
| Coimbatore | 5 | 25 | 30 |
| Tirupur | 5 | 25 | 30 |
The image features a bar chart illustrating dwell time in minutes
for several locations. The horizontal axis lists locations: Periyakulam,
Theni, Cumbum, Erode, Dindigul, Salem, Palani, Coimbatore, and Tirupur.
The vertical axis represents dwell time in increments up to one hundred
forty minutes, with Theni having the highest value of one hundred twenty
minutes. Dindigul follows at ninety minutes, along with Palani also at
ninety minutes. Other locations have lower values, with Salem at
forty-eight, and both Coimbatore and Tirupur at thirty minutes each. The
bars are black, clearly demarcating each location's respective dwell
time.LT of individual routes
The image features a bar chart illustrating dwell time in minutes
for several locations. The horizontal axis lists locations: Periyakulam,
Theni, Cumbum, Erode, Dindigul, Salem, Palani, Coimbatore, and Tirupur.
The vertical axis represents dwell time in increments up to one hundred
forty minutes, with Theni having the highest value of one hundred twenty
minutes. Dindigul follows at ninety minutes, along with Palani also at
ninety minutes. Other locations have lower values, with Salem at
forty-eight, and both Coimbatore and Tirupur at thirty minutes each. The
bars are black, clearly demarcating each location's respective dwell
time.LT of individual routes
6. Data analysis
The normalised matrix was used as the input for performing PCA to rank the influencing factors. Subsequently, queuing analysis was carried out to determine the capacity of the station. In addition, the conflicts inside the station locations are analysed using density analysis.
6.1 Principal component analysis
PCA performs the ranking of factors by reducing the dimensions of the responses received from the stakeholders (Fan et al., 2024; Hou et al., 2024; Rehan, 2021). The eigen analysis of the correlation matrix presented in Tables 9 and 10 shows the eigen vectors for the principal components. Figure 5 shows the scree plot of the principal components, whereas Figures 6 and 7 illustrate the biplot and loading plot of the principal components, respectively.
Eigen analysis of the correlation matrix
| Parameters | 2019 | 2021 | 2022 | 2023 | 2024 |
|---|---|---|---|---|---|
| Eigen value | 3.2046 | 1.0224 | 0.4562 | 0.1896 | 0.1272 |
| Proportion | 0.641 | 0.204 | 0.091 | 0.038 | 0.025 |
| Cumulative | 0.641 | 0.845 | 0.937 | 0.975 | 1.000 |
Eigen vectors
| Variable | PC1 | PC2 | PC3 | PC4 | PC5 |
|---|---|---|---|---|---|
| 2019 | 0.519 | −0.081 | 0.232 | 0.580 | −0.579 |
| 2021 | 0.475 | −0.402 | −0.292 | 0.297 | 0.663 |
| 2022 | 0.488 | −0.326 | −0.204 | −0.717 | −0.317 |
| 2023 | 0.429 | 0.412 | 0.680 | −0.246 | 0.353 |
| 2024 | 0.289 | 0.745 | −0.598 | 0.045 | −0.039 |
The horizontal axis represents component number from 1 to 5, and
the vertical axis represents eigenvalue. The curve shows a steep decline
from an eigenvalue of approximately 3.2 at component 1 to about 1.0 at
component 2, followed by a gradual decrease toward 0.1 by component 5.
The sharp drop between the first two components indicates that the first
component explains most of the variance, while subsequent components
contribute marginally. This pattern confirms that dimensionality
reduction beyond the first two components captures limited additional
information.Scree plot
The horizontal axis represents component number from 1 to 5, and
the vertical axis represents eigenvalue. The curve shows a steep decline
from an eigenvalue of approximately 3.2 at component 1 to about 1.0 at
component 2, followed by a gradual decrease toward 0.1 by component 5.
The sharp drop between the first two components indicates that the first
component explains most of the variance, while subsequent components
contribute marginally. This pattern confirms that dimensionality
reduction beyond the first two components captures limited additional
information.Scree plot
The horizontal axis represents the first component, and the
vertical axis represents the second component. Data points for the years
2019 to 2024 are plotted and connected with vectors emerging from the
origin, indicating how the components change across years. The vectors
show 2024 positioned farthest along the positive direction of the first
component, followed by 2023, 2019, 2022, and 2021. The overall spread of
points suggests gradual progression in both components over time, with
2024 showing the largest positive correlation among the observed
years.Biplot of principal components
The horizontal axis represents the first component, and the
vertical axis represents the second component. Data points for the years
2019 to 2024 are plotted and connected with vectors emerging from the
origin, indicating how the components change across years. The vectors
show 2024 positioned farthest along the positive direction of the first
component, followed by 2023, 2019, 2022, and 2021. The overall spread of
points suggests gradual progression in both components over time, with
2024 showing the largest positive correlation among the observed
years.Biplot of principal components
The image displays a loading plot illustrating the relationship
between the first component and the second component, where the x-axis
represents the first component and the y-axis represents the second
component. The plot contains five red lines, each labelled with a year
from 2019 to 2024, indicating their respective positions on the graph.
The year 2024 is positioned highest on the second component axis, while
2019 is located near the origin. The data points are plotted between
negative zero point five and positive zero point seventy-five on the
second component and from zero to zero point five on the first
component. The layout of the axes is clear, with the labels oriented for
easy reading.Loading plot of principal components
The image displays a loading plot illustrating the relationship
between the first component and the second component, where the x-axis
represents the first component and the y-axis represents the second
component. The plot contains five red lines, each labelled with a year
from 2019 to 2024, indicating their respective positions on the graph.
The year 2024 is positioned highest on the second component axis, while
2019 is located near the origin. The data points are plotted between
negative zero point five and positive zero point seventy-five on the
second component and from zero to zero point five on the first
component. The layout of the axes is clear, with the labels oriented for
easy reading.Loading plot of principal components
Table 11 lists the ranking of factors based on the response intensity. Based on the ranking analysis, the factors F1, F2, F3, F5, F6, F7 and F11 emerged as the most significant in influencing bus station design. Among these, Factor F11 is particularly critical, as it exhibits a zero correlation, value despite having a high average normalised score, highlighting its external importance, especially in terms of feeder service provision. In addition, Factors F1 and F2 are identified as high-priority considerations due to their relatively low normalised scores combined with significant correlation values, indicating their strong influence on the principal components and overall system performance.
Ranking
| Rank | Factor | Average normalised score | Correlation value |
|---|---|---|---|
| 1 | F11 | −1.0494 | 0.0000 |
| 2 | F1 | −0.9065 | 0.1429 |
| 3 | F2 | −0.9065 | 0.1429 |
| 4 | F7 | −0.5167 | 0.5327 |
| 5 | F6 | −0.2945 | 0.7549 |
| 6 | F5 | −0.2945 | 0.7549 |
| 7 | F3 | −0.2945 | 0.7549 |
| 8 | F12 | 0.2860 | 1.3353 |
| 9 | F4 | 0.8274 | 1.8768 |
| 10 | F8 | 1.0497 | 2.0991 |
| 11 | F9 | 1.0497 | 2.0991 |
| 12 | F10 | 1.0497 | 2.0991 |
Figure 8 presents the scoring and ranking of various factors influencing bus station design. The analysis reveals that factors such as F3, F5, F6, F7, F2, F1 and F11 require greater consideration in future bus station planning. Notably, Factor F11 exhibits a correlation value of 0.000, indicating no contribution to the principal component. This implies that while the feeder service (F11) may not directly influence internal station operations, its absence significantly affects traffic flow on the adjacent roads. Therefore, integrating feeder services is crucial for ensuring seamless external connectivity and reducing traffic disruptions around the bus station.
A bar chart illustrating the Principal Component Analysis (P C A)
scores of different factors, ranging from F eight to F twelve. The
vertical axis represents the PCA score, with values ranging from
negative one to positive one. The horizontal axis lists factors F eight
through F twelve. Bars corresponding to ranks two, four, five, six,
seven, nine, ten, eleven, and twelve are varied in height and colour,
with higher scores represented above zero and lower scores below zero.
The configuration indicates that ranks two through five have relatively
high scores, while rank twelve has the lowest score, positioned furthest
to the right.Ranking of factors
A bar chart illustrating the Principal Component Analysis (P C A)
scores of different factors, ranging from F eight to F twelve. The
vertical axis represents the PCA score, with values ranging from
negative one to positive one. The horizontal axis lists factors F eight
through F twelve. Bars corresponding to ranks two, four, five, six,
seven, nine, ten, eleven, and twelve are varied in height and colour,
with higher scores represented above zero and lower scores below zero.
The configuration indicates that ranks two through five have relatively
high scores, while rank twelve has the lowest score, positioned furthest
to the right.Ranking of factors
6.2 Queuing analysis
The study used queuing theory model, since both intercity and city services follow the random arrival of buses (Alam et al., 2021; An et al., 2021). The queue formation is calculated when the berth services in the station reaches its service capacity value. System state, i.e. number of buses in queuing system including buses in service bays, was considered.
The model follows two methods for calculating queuing system: one is M/M/1 system for single berth services for particular route of services and M/M/N for multichannel route services. The process of queuing starts with the bus arrival rate and passenger services, i.e. boarding/alighting, formation of queue and integration of services at bays. From the traffic observations, the number of buses arrival during peak hour is 160 buses/h. The actual capacity of the station can accommodate 90 buses per hour in all the ten routes at the same time with service intensity of 77 buses per hour.
Queuing of bus operations follows FIFO order. The basic function involves arrivals (arrival rate of buses), queuing (capacity and queue discipline) and service (number of bays). The service intensity was observed for peak hour traffic conditions. Since the buses follow random arrival rate, Poisson distribution formula can be used. Table 12 shows the arrival and service rate of buses per hour.
AR and SR of buses per hour
| R | n | At | St | SR | SR + queue/h (brc) |
|---|---|---|---|---|---|
| R1 | 1 | 11 | 10 | 7 | 13 |
| R2 | 1 | 12 | 20 | 6 | 12 |
| R3 | 1 | 9 | 10 | 6 | 12 |
| R4 | 1 | 13 | 9 | 8 | 14 |
| R5 | 1 | 9 | 15 | 5 | 11 |
| R6 | 1 | 14 | 8 | 6 | 12 |
| R7 | 1 | 11 | 15 | 5 | 11 |
| R8 | 1 | 16 | 5 | 12 | 18 |
| R9 | 1 | 15 | 5 | 10 | 16 |
Queuing value
| Route | ρ = (S = t) | P0 (1 − ρ) | Ls | Lq | Wq | WS | Q | Queue formation |
|---|---|---|---|---|---|---|---|---|
| R1 | 0.85 | 0.15 | 5.66 | 4.81 | 0.44 | 0.51 | −1.19 | No |
| R2 | 1 | Queue rate = maximum queue capacity | 0 | No | ||||
| R3 | 0.75 | 0.25 | 3 | 2.25 | 0.25 | 0.33 | −3.75 | No |
| R4 | 0.93 | 0.07 | 13.28 | 12.35 | 0.95 | 1.02 | 6.35 | Yes |
| R5 | 0.81 | 0.19 | 4.26 | 3.45 | 0.38 | 0.47 | −2.55 | No |
| R6 | 1.16 | Arrival rate > Service rate, i.e. Queue rate > Maximum queue capacity | 10.24 | Yes | ||||
| R7 | 1 | Queue rate = Maximum queue capacity | 0 | No | ||||
| R8 | 0.88 | 0.12 | 7.33 | 6.45 | 0.40 | 0.45 | 0.45 | Yes |
| R9 | 0.94 | 0.06 | 15.67 | 14.73 | 0.98 | 1.04 | 8.73 | Yes |
| Remarks: Zero in ‘Q’ indicates that no queue formation up to the value and negative sign indicates that no queue formed in the particular route | ||||||||
The bar graph presents three categories of data over nine time
intervals, labelled one through nine along the horizontal axis. The blue
bars represent the arrival rate of buses in buses per hour, with values
of eleven for intervals one, two, and three; twelve for intervals four,
six, and nine; and thirteen for interval eight. The red bars indicate
the service rate plus queue buses per hour, with varying values
including fourteen in intervals four and eight, while decreasing to zero
in interval seven. The green bars depict the queue numbers, showing
decreasing and varying values with the highest queue at eighteen in
interval eight and the lowest at zero in interval seven. The vertical
axis measures rates and queue numbers, while the horizontal shows time
intervals. Each category is distinctly coloured to differentiate the
data clearly, with numerical values displayed atop each bar.Queuing rate of bus routes
The bar graph presents three categories of data over nine time
intervals, labelled one through nine along the horizontal axis. The blue
bars represent the arrival rate of buses in buses per hour, with values
of eleven for intervals one, two, and three; twelve for intervals four,
six, and nine; and thirteen for interval eight. The red bars indicate
the service rate plus queue buses per hour, with varying values
including fourteen in intervals four and eight, while decreasing to zero
in interval seven. The green bars depict the queue numbers, showing
decreasing and varying values with the highest queue at eighteen in
interval eight and the lowest at zero in interval seven. The vertical
axis measures rates and queue numbers, while the horizontal shows time
intervals. Each category is distinctly coloured to differentiate the
data clearly, with numerical values displayed atop each bar.Queuing rate of bus routes
6.3 Congestion analysis
Congestion within the driveway area of the bus station was assessed using Q-GIS software to map the intensity of passenger movement (Gupta et al., 2025; Singh and Rangnekar, 2025; Pereira et al., 2024; Salazar-Carrillo et al., 2021). The analysis revealed that such congestion increases the risk of accidents, hampers bus services and compromises passenger safety.
Figure 10 illustrates the root causes contributing to this congestion, with the following key observations:
Passenger facilities are poorly located within the station, resulting in extended walking distances and encouraging unnecessary pedestrian movement within bus circulation zones.
As evident from the figure, nearly all areas within the terminal experience significant congestion, leading to frequent pedestrian-vehicle conflicts, service delay and occasional accidents.
The traffic survey identified that the terminal lacks proper turning provisions at entry and exit points. This forces buses to take wider turns, resulting in blockages at the front and delaying the entry of subsequent buses. This issue also extends to the adjacent carriageway, with an average delay time of ≈10 s.
Commercial shops situated at the far end of the terminal are not easily accessible to passengers.
The absence of dedicated pedestrian infrastructure contributes to unregulated and unsafe passenger movements across the terminal.
The plan shows the layout of the Madurai Arapalayam bus terminus,
with building areas and toilet zones outlined. Green rectangular shapes
represent buses, and the heat density indicates areas of varying
congestion intensity. The entrance and exit points are clearly marked
along with key facilities. High congestion density appears near the
central and southern portions, while the peripheral areas show
relatively less congestion. The map, created using Q G I S, visually
identifies traffic concentration zones useful for spatial design and
terminal management planning.Congestion mapping
The plan shows the layout of the Madurai Arapalayam bus terminus,
with building areas and toilet zones outlined. Green rectangular shapes
represent buses, and the heat density indicates areas of varying
congestion intensity. The entrance and exit points are clearly marked
along with key facilities. High congestion density appears near the
central and southern portions, while the peripheral areas show
relatively less congestion. The map, created using Q G I S, visually
identifies traffic concentration zones useful for spatial design and
terminal management planning.Congestion mapping
7. Design of layout
The proposed layout of the bus station with regulated traffic movement is shown in Figure 11. It minimises the generation of conflict points and service delays (Zhu et al., 2025). The planning steps include identifying the location, area requirements and dimensions of the outer boundary, followed by calculating the utility rate per hour for passengers and vehicles, and determining the traffic handling capacity for peak hour traffic. Next, the number of bays for each route is fixed, considering minimum layover times, while traffic predictions are made for design years. Area distribution for the site, including open and building areas, is determined, followed by the development of a land-use plan based on space requirements, bus flow and operations. An essential station components such as platforms, bays, terminal buildings and amenities are then incorporated, with pedestrian facilities designed according to IRC (2012) guidelines.
The image presents a layout map of a bus terminal. At the top of the
map is the terminal entry, leading to areas labeled as ticket counter,
dormitory, passenger utility area, emergency control room, and toilet on the
left side. The centre features multiple passenger platforms and bus
utility/circulation areas with thick outlines. Below these are additional
passenger utility areas and a series of shops arranged horizontally. At the
bottom, designated as the bus depot/maintenance area, are further passenger
utility areas and toilets, with clear markings for bus entries and exits
indicated by arrows along the sides. The layout allows for easy navigation
from the terminal entry through various sections towards bus
exits.Proposed bus station layout design
The image presents a layout map of a bus terminal. At the top of the
map is the terminal entry, leading to areas labeled as ticket counter,
dormitory, passenger utility area, emergency control room, and toilet on the
left side. The centre features multiple passenger platforms and bus
utility/circulation areas with thick outlines. Below these are additional
passenger utility areas and a series of shops arranged horizontally. At the
bottom, designated as the bus depot/maintenance area, are further passenger
utility areas and toilets, with clear markings for bus entries and exits
indicated by arrows along the sides. The layout allows for easy navigation
from the terminal entry through various sections towards bus
exits.Proposed bus station layout design
The traffic movements in the proposed layout are shown in Figure 12. The layout ensures seamless pedestrian flow, with specified widths for footpaths and walkways, acceptable walking distances within the transit area and strategically placed crosswalks. Traffic management measures utilising ITS technology (Parihar et al., 2022) are integrated to ensure smooth passenger flow from entry to platforms. The advantages of the layout proposed in the study include:
Efficient passenger flow: The design facilitates a streamlined movement of passengers.
Minimised walking distance: The layout reduces the walking distance within the bus station.
Accessible design: The station components are designed for accessibility, ensuring ease of use during layout preparation.
Versatile design: The layout is suitable for all types of bus station traffic intensity.
Enhanced accessibility for disabled individuals.
Reduced conflict points: The design significantly decreases potential conflict points.
The image presents a layout diagram for a bus terminal and
maintenance area. It includes several distinct sections such as passenger
utility areas labelled one, two, four, and five, as well as multiple
passenger platforms marked with yellow. The bus depot section is at the
bottom, featuring bus exits clearly indicated on both sides. Terminal entry
and exit points are marked at the top for navigation. The arrangement is
structured to facilitate movement with a clear flow for buses and passengers
through various designated areas. Visual markers represent different utility
zones and flow routes, ensuring a comprehensive overview of the terminal's
layout.Passenger and vehicle movement in the proposed layout
The image presents a layout diagram for a bus terminal and
maintenance area. It includes several distinct sections such as passenger
utility areas labelled one, two, four, and five, as well as multiple
passenger platforms marked with yellow. The bus depot section is at the
bottom, featuring bus exits clearly indicated on both sides. Terminal entry
and exit points are marked at the top for navigation. The arrangement is
structured to facilitate movement with a clear flow for buses and passengers
through various designated areas. Visual markers represent different utility
zones and flow routes, ensuring a comprehensive overview of the terminal's
layout.Passenger and vehicle movement in the proposed layout
7.1 Design specifications of the layout
The infrastructural requirements are designed per NBC and IRC specifications, including clearances, ITS provisions and security measures. Parking demand is assessed in alignment with IRC (2015), ensuring prioritised parking for public transportation, as well as strategies for congestion reduction and smart parking management. Lastly, landscaping and vegetation are incorporated to enhance the station’s visual appeal and reduce noise, with green areas strategically placed at the station’s front and entry, as per IRC (2009), along with recommended tree and turf specifications for footpaths and medians. Table 14 presents the space dimensions for the road adjacent to the bus station.
Space dimensions for adjacent roads
| Type of location | Spacing requirements | ||
|---|---|---|---|
| No. of lanes with divided carriage way | Exclusive bus lane length | ||
| Entry: m | Exit: m | ||
| CBD | 3-lane of width 3.75 m on each lane | 20 | 20 |
| Urban/Metropolitan areas | 3-lane of width 3.75 m on each lane | 20 | 10 |
Gandhi et al. (2015) provide the information about the space required for each components in bus station. Based on typology, average layover time, site area, traffic operations and bus flow per hour, the chart was developed. Since the study area serves for long route buses with fixed timings, the site area requirement is listed in Table 15. Similarly, Tables 15–20 show the different spatial requirements for various traffic operations.
Site area requirement
| Design period: 50 years, interstate bus terminal with fixed route bays Condition: At-grade parking with real estate development | ||||||
|---|---|---|---|---|---|---|
| Items | Vehicle | Chart (Pg. no.) | Average LT | BF/h | Area | Parking lot required |
| SAR | Bus | 116 | 60 | 300 | 34.58 acres | — |
| Car | — | 60 | — | — | 500 | |
Bay requirement
| Items | Vehicle | Chart (Pg. no.) | Type of bay | BF/h | No. of bays |
|---|---|---|---|---|---|
| Bay requirement | Bus | 120 | Boarding | 300 | 90 |
| — | — | Unloading | — | 15 | |
| — | — | Ideal | — | 275 |
Passenger accumulation, flow, floor area ratio and total built-up area requirement
| Items | Chart (Pg. no.) | Chart name | Requirement |
|---|---|---|---|
| PA | 122 | BF vs PA | 5000 |
| PF | 122 | BF vs PF | 13500 |
| FAR | 124 | BF vs FAR | 0.4 |
| Total built-up area | 126 | BF vs Built-up area | 65 000 m2 |
Site area distribution as per chart (Pg. no. 132) with LT of 60 min
| Items | Requirement | Area: acres |
|---|---|---|
| Bus work shop | 26.6% | 9.19 |
| Private vehicle parking | 27.4% | 9.47 |
| Feeder parking area | 11.6% | 4.01 |
| Circulation and pedestrian plaza | 21.6% | 7.46 |
| Terminal building | 12.8% | 4.14 |
Open area distribution as per chart (Pg. no. 131) for LT of 60 min
| Items | Requirement | Area: acres |
|---|---|---|
| Drop off | 6.5% | 0.26 |
| Pick up | 6.8% | 0.27 |
| Car parking | 31.4% | 1.26 |
| Bus parking | 29.9% | 1.20 |
| Circulation area | 24.8% | 1.00 |
| Workshop area | 0.6% | 0.02 |
At-grade bus parking requirement as per chart (Pg. no. 160)
| Items | Requirement | Area: acres |
|---|---|---|
| Boarding bay | 35.5% | 2.65 |
| Off-loading bay | 9.40% | 0.70 |
| Ideal bus bay | 55.1% | 4.11 |
As per Washington Metropolitan Area Transit Authority (2009), passenger boarding area of 1:50 sloping gradient with (2%) drainage is required. The study area is considered under medium-sized terminal (average flow – 300 buses/h) with average layover time of 60 min.
8. Conclusion
To conclude the research, the study highlights critical inefficiencies in the operational layout and service infrastructure of the Arapalayam Bus Station in Madurai. The increasing demand for bus transportation, coupled with inadequate facilities and poor integration with feeder services, has resulted in significant operational challenges including congestion, queuing delays and pedestrian-vehicle conflicts. The use of PCA helped identify key performance-limiting factors. To address these issues, the development of an integrated, multi-modal terminal with well-planned feeder services is essential. Such improvements would not only enhance traffic flow and passenger convenience but also promote a sustainable urban transport system. A well-planned layout could improve passenger convenience, reduce congestion and enhance safety. The highlights of the research includes:
Proposed an efficient ISBT layout with fixed bay allocation.
Designed to handle 3 lakh passengers daily, with 5% peak hour capacity.
Suitable for medium-scale terminals with 300 buses/h traffic flow.
Integrated feeder systems for city buses, taxis and private vehicles.
Achieved zero conflict points between vehicles and pedestrians.
Separated passenger zones from bus circulation to reduce conflicts and delay.
UN SDG’s related to the research work
UN SDG 9 – Industry, Innovation and Infrastructure
UN SDG 11 – Sustainable Cities and Communities
Suggestions for future work
The design is applicable to flat bus station type and the research regarding multi-level bus station design could be analysed.
Author’s contribution
Logeswaran S: conceptual design framework and site analysis; Anusha G: data collection and transportation analysis; Sampathkumar V: GIS analysis and spatial design; Anandakumar S: data collection and passenger movement analysis.
