Hubbry Logo
Bus bunchingBus bunchingMain
Open search
Bus bunching
Community hub
Bus bunching
logo
8 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Bus bunching
Bus bunching
Bus bunching
View on Wikipedia
from Wikipedia

Two buses together on the same route
Graphs of distance s vs time t illustrating bus-bunching:
1 Ideal journey
2 Bus B delayed by traffic congestion
3 B delayed by picking up passengers meant for C
4 C is early as B has picked up its passengers
5 Bus-bunching

In public transport, bus bunching, clumping, convoying, piggybacking or platooning is a phenomenon whereby two or more transit vehicles (such as buses or trains) that were scheduled at regular intervals along a common route instead bunch together and form a platoon. That occurs when the leading vehicles are unable to keep their schedule and fall behind to such an extent that trailing vehicles catch up to them.

Description

[edit]

A bus that is running slightly late will, in addition to its normal load, pick up passengers who would have taken the next bus if the first bus had not been late. These extra passengers delay the first bus even further. In contrast, the bus behind the late bus has a lighter passenger load than it otherwise would have had and may therefore run ahead of schedule. The classical theory causal model for irregular intervals is based on the observation that a late bus tends to get later and later as it completes its run, while the bus following it tends to get earlier and earlier.

Eventually, both buses form a pair, one right after the other, and the service deteriorates as the headway degrades from its nominal value. The buses that are stuck together are called a bus bunch or banana bus and may also involve more than two buses. The effect is often theorised to be the primary cause of reliability problems on bus and metro systems. Simulation studies have successfully demonstrated the extent of possible factors influencing bus bunching, and they may also be used to understand the impact of actions taken to overcome negative effects of bunching.[1]

Clumping can be caused by random heavy usage of any particular vehicle, which results in it falling behind schedule. The leading vehicle eventually lapses towards the time slot of a later scheduled vehicle. Sometimes, the later scheduled vehicle gets ahead of its own timetable, and both vehicles meet between their scheduled times. Sometimes, one scheduled vehicle may pass another.[2][3][4]

Clumping can be prevented or reduced as follows:

  • Scheduling minimum and maximum amounts of time at each stop[3]
  • Scheduling some crowded runs to skip certain stops[2]
  • If, on a popular route with frequent service, a crowded vehicle arrives, passengers can be urged to wait for the next vehicle, which may be less crowded.[2]

A different approach is to abandon the idea of a schedule and keep buses equally spaced by strategically delaying them at designated stops.[5] That is used to control the buses on the campus of Northern Arizona University, which outperform the previously scheduled system.[6]

A queueing theory paper in 1984 on multiple server cyclic queues observed the bus bunching effects and proposed a method called "dispersive schedules" to alleviate them.[7] Merely adding more vehicles to the schedule without making other changes has been proven not to be a reliable solution to the problem of bunching.[3]

See also

[edit]

References

[edit]
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

Bus bunching

View on Grokipedia
from Grokipedia
Bus bunching is a prevalent in public transit systems where buses operating on the same route arrive at stops in irregular clusters, often with multiple following closely together while leaving extended gaps without service, thereby deviating from scheduled headways. This clustering typically occurs when headways between consecutive buses fall below a threshold, such as 2 minutes, leading to uneven passenger loads and service disruptions. The root cause of bus bunching stems from a self-reinforcing loop involving headways, passenger demand, and dwell times: a bus that is delayed at one stop picks up more waiting passengers, extending its dwell time and further delaying it, while the subsequent bus encounters fewer passengers, shortens its dwell, and catches up to form a bunch. This inherent instability was first theoretically described in 1964 by Newell and Potts, who modeled bus propagation assuming constant ratios between passenger arrivals and boardings, uniform travel times between stops, and variability in en route, dwell, and holding durations, demonstrating how even small perturbations propagate and amplify along the route. Contributing factors include high passenger demand exceeding critical thresholds, traffic variability, signal timings, route length, and differences in driver speeds, which can induce synchronization-like bunching during peak periods. Bus bunching significantly impacts by increasing average passenger waiting times—often by 0.5 to 1.5 minutes per stop during incidents—causing on leading buses (with up to 10 additional passengers) and underloading of trailing ones, which erodes system efficiency and reliability. It also heightens operational challenges for transit agencies, such as elevated fuel consumption, emissions, and costs from inefficient vehicle utilization, while diminishing overall attractiveness. Empirical studies show bunching is most acute on high-frequency routes during rush hours, with initial irregularities at upstream stops propagating downstream and dissipating only partially at time-point controls. To counteract bus bunching, transit operators employ strategies like dynamic holding at stops to enforce even s, real-time crowding (RTCI) at stops to guide passengers toward less-loaded buses—potentially reducing headway variance by 40% and improving journey times by 6%—and advanced technologies such as GPS-based or inter-vehicle communication for . Emerging approaches, including synchronization modeling via coupled oscillators and hierarchical , further aim to predict and prevent clustering by addressing demand fluctuations and passenger swapping behaviors.

Overview

Definition

Bus bunching is a observed in public transit systems, particularly on fixed-route bus lines, where multiple vehicles arrive at stops in irregular clusters rather than at consistent intervals, resulting in simultaneous or near-simultaneous arrivals followed by prolonged service gaps. This leads to uneven distribution of passengers and reduced . The term encompasses similar issues in other transit modes, such as trains or trams, but is most commonly associated with buses. At its core, bus bunching arises from instability in , defined as the time or spatial interval between consecutive vehicles on a route. Small initial deviations in headway—such as a bus departing a stop slightly later than scheduled—can amplify progressively along the route due to inherent dynamics in transit operations, causing vehicles to converge into groups known as platoons. Headway variance, which quantifies the irregularity in these intervals, serves as a key metric for assessing the severity of bunching, with higher variance indicating greater clustering. This instability was first systematically analyzed in seminal work by Newell and Potts (1964). A fundamental mechanism driving this process is a in vehicle operations: a delayed bus accumulates more waiting passengers at subsequent stops, extending its dwell time (the period spent at a stop for boarding and alighting), while the trailing bus encounters fewer passengers and shorter dwell times, enabling it to close the gap and bunch up. This self-reinforcing cycle promotes platoon formation, where clustered buses effectively operate as a single unit with shared passenger loads. The can be visually illustrated through simple diagrams contrasting ideal even headways—depicting buses arriving at regular intervals—with bunched scenarios, where multiple buses converge at a stop amid extended voids, highlighting the impact on service uniformity.

Historical Context

The of bus bunching, characterized by the clustering of vehicles on a route leading to irregular headways, was first systematically analyzed in the mid-20th century amid growing concerns over urban transit reliability. A key milestone came in 1964 with the seminal work by G. F. Newell and R. B. Potts, who developed a foundational deterministic model illustrating how initial delays propagate through a bus line, causing trailing vehicles to catch up to leading ones and form bunches. This model, presented in their paper "Maintaining a Bus Schedule," assumed uniform passenger arrivals and constant travel times between stops, providing the first theoretical explanation for the instability inherent in high-frequency bus operations. Subsequent developments in the and advanced this through , emphasizing headway-based control strategies to stabilize schedules, such as vehicle holding and speed adjustments. Research evolved significantly in the 1980s toward stochastic models that incorporated randomness from variable travel times, passenger boarding variability, and traffic conditions, offering more realistic simulations of bunching dynamics. For instance, approaches captured the probabilistic nature of bus movements along routes, highlighting how small perturbations amplify into widespread irregularities. From the late 1990s onward, the proliferation of automatic vehicle location (AVL) systems facilitated data-driven empirical studies, intensifying focus on bunching due to rapid urban growth and increased transit demand in megacities. These technologies enabled real-time analysis of deviations, revealing bunching's prevalence in operational data from various networks. Notable examples underscore the persistence of bunching despite early interventions. In during the late 1970s and 1980s, transit authorities grappled with frequent bunching on busy routes, employing two-way radios to instruct drivers on slowing or holding to restore spacing. These cases illustrated bunching's endurance as a challenge, even as technological and analytical advances progressed.

Causes

Operational Factors

One key operational factor contributing to bus bunching is dwell time variability at stops, where delayed buses accumulate more passengers due to longer wait times for subsequent arrivals, leading to extended boarding and alighting durations that further exacerbate delays in a loop. This phenomenon is particularly pronounced on high-frequency routes, as passenger demand concentrates on later vehicles, increasing their dwell times relative to on-schedule buses ahead. Studies using automatic passenger counting data have shown that such variability can account for a significant portion of overall instability. Dispatch and scheduling issues also initiate and compound bus bunching by introducing uneven initial headways from depot departures or inadequate recovery times at terminals. For instance, if buses are dispatched with irregular intervals—often due to manual processes or insufficient slack in timetables—small deviations propagate downstream, as trailing buses catch up to leaders without corrective measures. on transit operations indicates that optimizing dispatch headways through real-time monitoring can reduce initial irregularities, preventing the cascading effects that lead to platoons. Vehicle interactions, such as platooning at stops, further amplify bunching by limiting opportunities and forcing following buses to idle behind leaders, which diminishes speed differentials and compresses s. In urban settings with curb-side stops, this dynamic reduces the ability of faster buses to regain lost time, as queues form and effective travel speeds converge across the fleet. Simulations incorporating these interactions demonstrate that without dedicated bus lanes or provisions, variances can increase over a route due to such constraints. Driver behavior plays a subtle yet critical role in exacerbating small delays that contribute to bunching, through variations in acceleration and deceleration patterns or inconsistent adherence to schedules. For example, drivers may accelerate to recover time after or decelerate excessively at stops, introducing variability in running times that accumulates across vehicles. Empirical analyses of automatic vehicle location data reveal that these behavioral factors contribute to deviations, particularly when drivers respond reactively to real-time schedule information without standardized protocols.

External Influences

Traffic significantly contributes to bus bunching by introducing variability in travel speeds along routes shared with other vehicles. In mixed traffic environments, buses experience delays from fluctuating road conditions, where a leading bus slowed by congestion accumulates more waiting passengers at subsequent stops, extending its dwell time and further delaying it. Meanwhile, a trailing bus, encountering fewer passengers due to the longer , maintains higher speeds and catches up, initiating or exacerbating bunching. This dynamic is amplified during periods of high roadway demand, where reduced speeds create shock waves that propagate backward, forcing vehicles—including buses—into closer formations and irregular spacing. Fixed or poorly coordinated traffic signal timings at intersections also promote bus bunching by disproportionately affecting leading vehicles. An unexpected red signal can impose additional on the first bus in a , increasing its dwell time at the next stop and allowing subsequent buses to arrive sooner relative to the . Without adaptive signal priority, the probability of such correlates with the red-to-cycle length ratio, leading to instability as trailing buses proceed through green phases more readily. Simulations indicate that human-driven buses exhibit bunching under fixed signal cycles, highlighting the role of uncoordinated intersections in perpetuating this issue. Uneven demand patterns, such as peak-hour surges or Poisson-distributed arrivals at stops, overload specific buses and intensify bunching through extended dwell times. During rush periods, sudden influxes of passengers cause the leading bus to spend longer boarding, widening and prompting more arrivals for the next vehicle, which then accelerates to catch up. models demonstrate that non-uniform arrival patterns significantly influence bunching probability, even absent other delays, as variability in disrupts headway regularity and amplifies interactions with external slowdowns. For instance, case studies show that insufficient boarding capacity under high demand leads to clustered arrivals, where the effects compound across stops. Weather conditions and incidents like accidents or further degrade route capacity, introducing unpredictable delays that foster bunching. Rain and snow reduce traffic speeds and encourage cautious boarding, elevating the likelihood of extreme lateness while also potentially causing earliness through fewer passengers and shorter dwells. Similarly, accidents generate localized congestion, delaying buses and causing trailing ones to cluster at downstream stops, as seen in analyses of travel time variability. Construction work narrows lanes or alters paths, mimicking these effects by stochastically impeding flow and allowing faster buses to close gaps on slowed leaders.

Consequences

Service Reliability

Bus bunching significantly undermines adherence in transit systems by amplifying headway variance as buses progress downstream. This results in a marked increase in the (CV) of headways, often reaching a of 0.59 across high-frequency urban routes, with ranges typically between 0.4 and 0.8. Such variability leads to schedule deviations that can exceed 50% in severe bunching scenarios, as disturbances propagate and exacerbate inconsistencies in bus spacing. For instance, terminal headway variability alone can increase CV by 0.373, directly contributing to unreliable service patterns observed in networks like Chicago's CTA. Operationally, bus bunching elevates costs through inefficient resource use and heightened consumption. Idling in bunches at stops causes unproductive halts, leading to loss ratios of 2.91% to 17.26% per trip in urban-rural corridors, as buses wait unnecessarily while gaps form elsewhere. Recovery efforts to address these disruptions often necessitate increased for drivers, further straining budgets, while underutilization during extended gaps reduces overall fleet . These inefficiencies compound as bunching forces operators to deploy additional recovery measures, diminishing the effective capacity of existing vehicles. At the system level, bunching propagates across routes and stops, reducing overall fleet capacity and compelling agencies to allocate more vehicles to sustain service frequencies. This propagation effect, where initial delays cascade through the network, can link up to 29% of bunching events to preceding gaps, thereby lowering the system's ability to handle demand without excess resources. Reliability indicators suffer accordingly, with on-time performance declining sharply; analyses of overlapping services show headway delays increase by up to 3.8 seconds per stop. This manifests in symptoms like uneven passenger waiting times, highlighting the broader dependability challenges.

Passenger Experience

Bus bunching significantly disrupts passenger waiting times by creating irregular gaps between services, where some buses arrive in close succession while others are delayed far beyond schedule. This variability leads to longer average waits, with empirical analyses indicating that the weighted average passenger waiting time for bunched bus trips can be approximately 0.5 minutes longer than for unbunched operations. In severe cases, passengers may endure waits far beyond the scheduled , as clusters of buses pass empty stops followed by extended voids, amplifying frustration and perceived unreliability. The phenomenon also results in overcrowding on bunched buses, where accumulated demand overloads vehicles, causing standing room shortages and reduced comfort levels. Passenger densities can reach up to 6 passengers per square meter during peak periods in affected systems, leading to heightened discomfort, longer in-vehicle times, and dissatisfaction with overall . Conversely, the preceding gaps leave bus stops deserted, forcing some riders to miss connections or abandon trips altogether, further compounding the uneven passenger experience. Bus bunching disproportionately impacts low-income and transit-dependent users, who comprise a of bus riders in many urban networks and often travel in high-bunching corridors with limited alternatives. This exacerbates access barriers to like and healthcare, as unreliable service hinders timely arrivals and increases the effective cost of transit use for these groups. surveys highlight how such inequities amplify broader mobility challenges for vulnerable populations reliant on buses. In response to these disruptions, passengers frequently report high levels of frustration, with surveys identifying perceived unreliability—often attributed directly to bus bunching—as a top in transit systems. This dissatisfaction can prompt behavioral shifts, such as abandoning buses for private cars or ridesharing services, potentially undermining public transit's goals. For instance, extended waits and have been linked to mode shifts away from buses, reducing overall ridership and perpetuating service instability.

Modeling

Deterministic Models

Deterministic models of bus bunching provide a theoretical foundation by assuming fixed parameters, such as constant travel times between stops and uniform boarding rates, to analyze the of and without incorporating . These models highlight the inherent in bus operations, where small initial perturbations can amplify over time, leading to clustering. The seminal Newell-Potts model, developed in 1964, describes the dynamics of in a simplified bus route scenario. It posits that deviations from scheduled arise primarily from differential dwell times at stops, influenced by the time since the previous bus arrived. The model demonstrates how larger result in proportionally longer dwell times for the following bus due to accumulated waiting s, causing growth in headway variance and inevitable bunching downstream. Under the model's assumptions of constant speeds between stops and no stochastic elements, it predicts bunch formation even in ideal conditions, as the feedback loop between and dwell time destabilizes the system. This approach has informed early simulations for route design, demonstrating that bunching emerges as a natural outcome without corrective measures. Fluid approximations extend these ideas by treating the bus fleet as a continuous flow rather than discrete vehicles, focusing on aggregate delay along the route. These models assume flow and constant parameters to trace how speed reductions—due to fixed factors like dwell or —accumulate, fostering formation. Applications include baseline predictions for , underscoring bunching's persistence in deterministic settings.

Stochastic Models

models of bus bunching extend deterministic frameworks by incorporating probabilistic elements to account for uncertainties in travel times, dwell times, and passenger demands, providing a more realistic representation of operational variability. These models treat bus s as random processes, enabling predictions of bunching probabilities and headway distributions under noisy conditions. Markov chain approaches model headways as Markov processes, where the headway state at successive stops evolves according to transition probabilities derived from stochastic factors such as variable travel times and dwell durations. Passenger arrivals at stops are often assumed to follow a Poisson process, leading to random boarding times that influence dwell variability and headway transitions. In these models, headway variance accumulates along the route due to additive noise, as captured by the recursive relation Var⁡(hn+1)=Var⁡(hn)+σ2\operatorname{Var}(h_{n+1}) = \operatorname{Var}(h_n) + \sigma^2, where σ2\sigma^2 quantifies the variance from random disturbances like traffic fluctuations or passenger loads. Simulation-based models, including simulations, leverage empirical data such as Automatic Vehicle Location (AVL) records to estimate bunching probabilities amid variable demand patterns. These approaches generate multiple scenarios of bus operations, incorporating random interactions to replicate real-world dynamics. For instance, agent-based simulations treat buses and passengers as autonomous agents with behaviors, such as random alighting and boarding, allowing analysis of how demand variability propagates bunching along a route. Recent advances incorporate techniques, such as decision tree ensembles and , to predict bus operation states and bunching risks using AVL and real-time data, enhancing forecast accuracy for control strategies as of 2023. Key extensions to these models address , such as differences in stop densities or route segments, which alter the likelihood of buses converging. Varying stop densities can increase bunching in denser areas by amplifying dwell time variability. The expected bunch size in such scenarios follows a , given by E[B]=11−pE[B] = \frac{1}{1 - p}, where pp represents the probability that a trailing bus catches up to the leading one due to speed adjustments or delays. Validation of stochastic models relies on empirical comparisons using AVL and passenger count data from operational systems. In studies from cities like and , these models accurately reproduce observed distributions and bunching events, with stochastic components explaining 30-60% of the total variance in headways compared to deterministic baselines.

Mitigation Strategies

Operational Adjustments

Operational adjustments to mitigate bus bunching involve procedural changes in transit operations, such as shifting scheduling paradigms, incorporating buffer times, enhancing driver protocols, and refining route configurations to promote even distribution without relying on advanced . These strategies aim to interrupt the feedback loops that amplify delays and clustering, drawing from established transit engineering principles. Headway-based scheduling represents a shift from fixed timetables to dynamic control focused on maintaining consistent intervals between buses. In this approach, operators monitor real-time and apply holding rules at designated control points to prevent deviations; for instance, a trailing bus is held if the observed headway falls below a tolerance threshold, such as 50% of the scheduled interval, allowing the leading bus to gain separation. This method outperforms traditional schedule adherence by requiring less overall slack, thereby reducing passenger waiting times and increasing , as demonstrated in systematic analyses of urban routes. Slack time insertion provides buffers at terminals or along routes to absorb delays, stabilizing departures and reducing initial variance that propagates bunching. Optimal slack ratios of 0.1 to 0.2—added to scheduled times—convexly decrease mean and variance of delays, ensuring stability under schedule-based control and minimizing expected passenger waits in multi-bus loops, according to queueing models and simulations. Driver emphasizes protocols to preserve spacing, including speed advisories that guide operators to adjust velocities based on distances to preceding and following buses, thereby avoiding inadvertent or lagging. Additional maneuvers, such as brief holds or selective skipping at low-demand stops, enable bunch-breaking without disrupting high-ridership areas; these guidelines, when integrated into routine , enhance regularity by interrupting clustering dynamics, as evidenced in reviews of operational interventions. allowances in designated zones further support even distribution, reducing waiting times in simulated bunching scenarios. Route design tweaks target dwell time variability by optimizing stop placements and service frequencies to align with demand patterns, minimizing boarding-induced delays that exacerbate bunching. Closer stop spacing in dense areas reduces alighting variability, while frequency adjustments—such as increasing service during peak hours—distribute load more evenly; case studies from Barcelona's , including route H10, illustrate these changes yielding 24% reductions in variation coefficients and improved adherence, achieving up to 40% gains in overall service regularity through refined timetables and branching configurations.

Technological Interventions

Technological interventions for mitigating bus bunching leverage collection, , and predictive algorithms to adjust bus operations dynamically, aiming to maintain even and reduce service variability. Automatic Vehicle Location (AVL) systems, which use GPS tracking to monitor bus positions, enable holding strategies where buses are paused at stops if their falls below a predefined threshold, such as 50-70% of the scheduled interval. This approach counteracts bunching by allowing trailing buses to catch up, with empirical studies demonstrating reductions in variance by up to 12.9% in headway-based operations. In Portland's Tri-Met system, AVL-integrated holding has contributed to a 4% decrease in headway variation alongside improvements in running time adherence. Transit Signal Priority (TSP) systems further address bunching by communicating bus locations to signals via AVL data, granting dynamic green extensions or phase insertions to approaching buses, thereby preserving momentum and minimizing stop-and-go delays. These interventions are particularly effective in congested urban corridors, where models and field tests indicate 10-20% reductions in bus delays; for instance, a in medium-congested scenarios showed significant delay cuts without substantially impacting cross-street . In practice, TSP implementations like those evaluated by the Virginia Transportation Research Council achieved up to 23% reductions in transit vehicle delays at signals. Predictive analytics powered by artificial intelligence (AI) and machine learning (ML) forecast bunching risks by analyzing historical AVL data, traffic patterns, and passenger loads, enabling proactive adjustments such as speed guidance or supplemental dispatching. Singapore's (LTA) integrates such tools into its bus management apps, using ML to predict arrival times and optimize en-route regularity, which has reduced bus bunching instances and improved by facilitating real-time interventions. These systems process vast datasets to identify deviations early, supporting centralized coordination that enhances overall network reliability. Emerging technologies, including vehicle-to-infrastructure (V2I) communication, enable bus platooning where connected buses maintain tight formations to stabilize s and reduce the propagation of delays. Post-2020 smart city pilots, such as those explored in automated concepts, demonstrate V2I's potential for coordinated control, allowing platoons to adjust speeds collectively and mitigate bunching in mixed traffic. For electric buses, battery optimization strategies address dwell variations caused by charging, with dynamic holding and speed guidance at wireless charging stops reducing bunching cycles; research on at-stop charging shows these methods improve charging efficiency while stabilizing s. Recent advancements as of 2024 include hierarchical (HMARL) frameworks that integrate holding and passenger information strategies to further reduce headway deviations, and modular bus designs that enhance bunching resistance by allowing flexible vehicle reconfiguration.

References

  1. https://pdxscholar.library.pdx.edu/cgi/viewcontent.cgi?article=1328&context=cengin_fac
  2. https://pmc.ncbi.nlm.nih.gov/articles/PMC8895103/
  3. https://digitalcommons.usf.edu/cgi/viewcontent.cgi?article=1906&context=jpt
  4. https://pmc.ncbi.nlm.nih.gov/articles/PMC6499774/
  5. https://rosap.ntl.bts.gov/view/dot/24701/dot_24701_DS1.pdf
  6. https://australasiantransportresearchforum.org.au/wp-content/uploads/2022/05/ATRF2022_Resubmission_77.pdf
  7. https://www.nature.com/articles/s41598-019-43310-7
  8. https://www.semanticscholar.org/paper/Maintaining-a-bus-schedule-Newell-Potts/8811ce56d4672109b2b02c2935425bfd3a67badb
  9. https://www.sciencedirect.com/science/article/abs/pii/0191260783900031
  10. https://onlinepubs.trb.org/onlinepubs/tcrp/tcrp_syn_24.pdf
  11. https://www.nytimes.com/1979/07/14/archives/letters-the-mtas-answer-to-bus-bunching-con-ed-is-ready-to-testburn.html
  12. https://comptroller.nyc.gov/reports/the-other-transit-crisis-how-to-improve-the-nyc-bus-system/
  13. https://onlinepubs.trb.org/onlinepubs/tcrp/docs/tcrp100/Part4.pdf
  14. https://journals.sagepub.com/doi/10.3141/2072-13
  15. https://www.researchgate.net/publication/245563041_Modeling_the_Factors_Affecting_Bus_Stop_Dwell_Time_Use_of_Automatic_Passenger_Counting_Automatic_Fare_Counting_and_Automatic_Vehicle_Location_Data
  16. https://www.sciencedirect.com/science/article/abs/pii/S0191261509000484
  17. https://www.researchgate.net/publication/252093865_Improving_Scheduling_Through_Performance_Monitoring_Using_AVLAPC_Data
  18. https://www.researchgate.net/publication/318333047_Modelling_bus_bunching_and_holding_control_with_vehicle_overtaking_and_distributed_passenger_boarding_behaviour
  19. http://cake.fiu.edu/Publications/Luckerath%2Bal-18-MB.A.Mesoscopic_Bus_Transit_Simulation_Model_SNE_downloaded.pdf
  20. https://pdxscholar.library.pdx.edu/cgi/viewcontent.cgi?article=1082&context=cus_pubs
  21. https://www.researchgate.net/publication/328655697_Determinants_of_Bus_Riding_Time_Deviations_Relationship_between_Driving_Patterns_and_Transit_Performance
  22. https://doi.org/10.1016/j.trb.2009.04.002
  23. https://ops.fhwa.dot.gov/congestion_report/chapter2.htm
  24. https://doi.org/10.5038/2375-0901.23.1.3
  25. https://www.researchgate.net/publication/278850034_A_Model_of_Bus_Bunching_under_Reliability-based_Passenger_Arrival_Patterns
  26. https://doi.org/10.1016/j.trc.2025.105000
  27. https://doi.org/10.3390/smartcities8050141
  28. https://www.tandfonline.com/doi/full/10.1080/01441647.2024.2313969
  29. https://doi.org/10.1371/journal.pone.0290903
  30. https://tram.mcgill.ca/Research/Publications/reliability_bunching.pdf
  31. https://www.researchgate.net/publication/354737641_Headway_variability_in_public_transport_a_review_of_metrics_determinants_effects_for_quality_of_service_and_control_strategies
  32. https://theconversation.com/math-explains-why-your-bus-route-seems-so-unreliable-95485
  33. https://publications.iadb.org/publications/english/document/Evaluation-of-Passenger-Comfort-in-Bus-Rapid-Transit-Systems.pdf
  34. https://www.urbanstudiesjournal.com/how-buses-alleviate-unemployment-and-poverty/
  35. https://www.ridegrtc.com/community/transit-access-partnership/
  36. https://www.sciencedirect.com/science/article/pii/S019126151730989X
  37. https://www.nickmalleson.co.uk/papers/2019-BusSimTransportmetrica.pdf
  38. https://caspt.org/wp-content/uploads/2018/10/Papers/CASPT_2018_paper_56.pdf
  39. https://escholarship.org/content/qt6zc5j8xg/qt6zc5j8xg.pdf
  40. https://www.researchgate.net/publication/327561567_Agent-Based_Approach_to_Modelling_Bus_Bunching_and_Dispersion_Under_the_Influence_of_Demand_Variability
  41. https://www.researchgate.net/publication/341679905_Modeling_and_Prediction_of_Bus_Operation_States_for_Bunching_Analysis
  42. https://www.sciencedirect.com/science/article/pii/S2046043020300289
  43. https://www.sciencedirect.com/science/article/pii/S0968090X22003667
  44. https://www.researchgate.net/publication/266584951_Empirical_findings_of_Bus_Bunching_Distributions_and_Attributes_Using_Archived_AVLAPC_Bus_Data
  45. https://dl.acm.org/doi/10.1155/2021/2606191
  46. https://pubsonline.informs.org/doi/10.1287/trsc.1060.0170
  47. https://doi.org/10.1007/s42524-024-0297-1
  48. https://upcommons.upc.edu/server/api/core/bitstreams/f281a67f-8788-4373-86b6-092b207be959/content
  49. https://attap.umd.edu/wp-content/uploads/2022/05/Transit_Priority_Strategies_for_Multiple_Routes_under_Headway_based_Operations.pdf
  50. https://trb.org/publications/tcrp/tcrp_webdoc_23.pdf
  51. https://www.sciencedirect.com/science/article/pii/S2046043023000035
  52. https://vtrc.virginia.gov/media/vtrc/vtrc-pdf/vtrc-pdf/06-cr6.pdf
  53. https://eprints.whiterose.ac.uk/106500/1/Leong_Goh_Hess_Murphy_RETREC_2016.pdf
  54. https://www.researchgate.net/publication/396097591_Singapore%27s_Land_Transport_Authority_LTA_A_Case_Study_of_Predictive_AI_and_Centralized_Coordination_in_Urban_Traffic_Management
  55. https://www.wsp.com/-/media/insights/us/bus-automation-whitepaper/bus-automation-whitepaper-wsp-2023.pdf
  56. https://www.tandfonline.com/doi/full/10.1080/0305215X.2025.2535526?src=
  57. https://ieeexplore.ieee.org/document/10440179/
  58. https://pubsonline.informs.org/doi/10.1287/trsc.2023.0224
Add your contribution
Related Hubs
User Avatar
No comments yet.