Modelling driver behaviour towards innovative warning devices at railway level crossings

Tey, Li-Sian; Wallis, Guy; Cloete, Steven; Ferreira, Luis

doi:10.1016/j.aap.2012.11.002

Cited by 31 publications

(16 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A recent application of fuzzy sets and clustering to guide the selection of rail crossings for active safety systems (e.g., bells, lights, and barriers) is in [3]. Tey et al [4] describe the use of logistic regression and mixed regression to model the behavior of drivers at railway crossings. The paper by Akin and Akbas [5] describes the use of neural networks to model intersection crashes and intersection characteristics, such as, lighting, surface materials, etc.…”

Section: Related Workmentioning

confidence: 99%

Text Mining the Contributors to Rail Accidents

Brown

2016

IEEE Trans. Intell. Transport. Syst.

View full text Add to dashboard Cite

Rail accidents represent an important safety concern for the transportation industry in many countries. In the 11 years from 2001 to 2012, the U.S. had more than 40 000 rail accidents that cost more than $45 million. While most of the accidents during this period had very little cost, about 5200 had damages in excess of $141 500. To better understand the contributors to these extreme accidents, the Federal Railroad Administration has required the railroads involved in accidents to submit reports that contain both fixed field entries and narratives that describe the characteristics of the accident. While a number of studies have looked at the fixed fields, none have done an extensive analysis of the narratives. This paper describes the use of text mining with a combination of techniques to automatically discover accident characteristics that can inform a better understanding of the contributors to the accidents. The study evaluates the efficacy of text mining of accident narratives by assessing predictive performance for the costs of extreme accidents. The results show that predictive accuracy for accident costs significantly improves through the use of features found by text mining and predictive accuracy further improves through the use of modern ensemble methods. Importantly, this study also shows through case examples how the findings from text mining of the narratives can improve understanding of the contributors to rail accidents in ways not possible through only fixed field analysis of the accident reports.Index Terms-Rail safety, safety engineering, latent Dirichlet allocation, partial least squares, random forests. 1524-9050

show abstract

Section: Related Workmentioning

confidence: 99%

Text Mining the Contributors to Rail Accidents

Brown

2016

IEEE Trans. Intell. Transport. Syst.

View full text Add to dashboard Cite

show abstract

“…For example, Tey et al compared driving behaviors from field data and a driving simulator for compliance rate, speed profile, and final breaking position at railway crossings equipped with a stop sign, flashing lights, and half boom barrier (11). In another study by Tey et al (12), four warning devices-flashing lights, in-vehicle warning, rumble strips, and stop sign-were tested according to the age and gender of the participants. Tey et al used a fixed driving simulator to identify compliance rate, driver accelerator release position, and initial and final breaking positions.…”

Section: Use Of Driving Simulator and Traffic Simulation For Railway mentioning

confidence: 99%

Traffic Safety at Road–Rail Level Crossings Using a Driving Simulator and Traffic Simulation

Kim

Larue

Ferreira

et al. 2015

Transportation Research Record: Journal of the Transportation R

Self Cite

View full text Add to dashboard Cite

Several intelligent transportation systems (ITS) were used with an advanced driving simulator to assess its influence on driving behavior. Three types of ITS interventions were tested: video in vehicle, audio in vehicle, and on-road flashing marker. The results from the driving simulator were inputs for a developed model that used traffic micro-simulation (VISSIM 5.4) to assess the safety interventions. Using a driving simulator, 58 participants were required to drive through active and passive crossings with and without an ITS device and in the presence or absence of an approaching train. The effect of changes in driver speed and compliance rate was greater at passive crossings than at active crossings. The slight difference in speed of drivers approaching ITS devices indicated that ITS helped drivers encounter crossings in a safer way. Since the traffic simulation was not able to replicate a dynamic speed change or a probability of stopping that varied depending on ITS safety devices, some modifications were made to the traffic simulation. The results showed that exposure to ITS devices at active crossings did not influence drivers' behavior significantly according to the traffic performance indicator, such as delay time, number of stops, speed, and stopped delay. However, the results of traffic simulation for passive crossings , where low traffic volumes and low train headway normally occur, showed that ITS devices improved overall traffic performance. In the United States, empirical formulas based on historical accident data at level crossings have been used to predict the expected crash rate. These formulas, such as the Peabody-Dimmick formula, the New Hampshire index, the NCHRP crash prediction formula, the U.S. Department of Transportation crash prediction formula, and the Mississippi and Ohio methods, consider the crash history as well as some of the causal factors in determining the crash rate at a particular crossing. While a hazard index is a relative ranking, the crash prediction models calculate the actual frequency of crashes at crossings (1). Statistical collision prediction models are used to assess how specific countermeasures act to reduce collisions at specific grade crossings. In Australia, the Australian level-crossing assessment model is used to identify contributing risk factors at level crossings. This tool can be used to prioritize the level crossings that are to be upgraded. Although the procedure for archiving crash data appears to have become more systematic, it often contains significant discrepancies. Many crash-related organizations, such as police, insurance companies , and bureaus of statistics, collect crash data in different ways. Police reports are prone to underreported bias. Elvik and Mysen analyzed crash recording rates in 13 countries (2). In their study, only 95% of fatal crashes, 70% of serious injury crashes (hospitalized), 25% of slight injury crashes (outpatients), 10% of very slight injuries (sent home), and 25% of property-damage-only crashes were reported, compared...

show abstract

“…Drivers tend to decelerate in a discontinuous manner when they are approaching HRGCs [22,23]. Drivers tend to reduce their speeds when they initially perceive risks.…”

Section: Demand Functionsmentioning

confidence: 99%

“…Only the single vehicle data from both sites are used in this study. The research of Tey et al [22] using an HRGC driving simulator provides detail distance statistics (means and standard errors) of accelerator release, initial braking, and final braking for two initial speeds (60 and 80 kph). The PDFs of all variables in the demand functions are defined in Table 2 [24].…”

Section: Data Sourcesmentioning

confidence: 99%

Transmission Range Evaluations for Connected Vehicles at Highway-Rail Grade Crossings

Hsu¹,

Jones

2017

Designs

View full text Add to dashboard Cite

This study evaluates the transmission range requirements of Connected Vehicles (CVs) at Highway-Rail Grade Crossings (HRGCs) in terms of safety improvement. The safety improvement of HRGCs is evaluated by using a reliability-based risk analysis that calculates risk of collision for CVs and non-CVs. Trains are assumed to have onboard units that transmit train location and speed information to CVs via vehicle to vehicle communications. The stopping distance and time to collision of a vehicle are the demand functions in reliability-based risk analysis. The demand functions consist of probability density functions of a vehicle's initial speed, perception-reaction time, initial deceleration rate, final speed, and final deceleration rate. Train arrival time depending on the train speed and transmission range is the supply threshold for calculating the CV's risk of collision at passive HRGCs. The transmission range's projected highway distance is the supply threshold for CVs at active HRGCs. After deriving probability density functions of demand functions from the published data, Monte Carlo simulations generate the probabilities or risks that a CV would fail to stop within the transmission range or train arrival time. With the provision of a 600 m transmission range, the risk of collision for the CV is lower than that for the non-CV with a 300 m sight distance to the train at the passive HRGC. The CV's risk of collision is lower than the non-CV's with a 300 m transmission range at active HRGCs. The CV application at HRGCs can improve safety by reducing CVs' risk of collision. A 600 m transmission range is desirable at passive HRGCs. A 300 m transmission is sufficient for CVs at active HRGCs. Overall, a 600 m transmission range is feasible to improve the safety at passive and active HRGCs.

show abstract

Modelling driver behaviour towards innovative warning devices at railway level crossings

Cited by 31 publications

References 10 publications

Text Mining the Contributors to Rail Accidents

Text Mining the Contributors to Rail Accidents

Traffic Safety at Road–Rail Level Crossings Using a Driving Simulator and Traffic Simulation

Transmission Range Evaluations for Connected Vehicles at Highway-Rail Grade Crossings

Contact Info

Product

Resources

About