Service quality analysis of signalized intersections from the perspective of bicycling

https://doi.org/10.1016/j.jth.2020.100827Get rights and content

Highlights

  • The bicycle service quality at signalized intersections is analyzed and modeled.

  • Functional network-based BLOS model has higher prediction ability than many others.

  • Construction of pedestrian crossover bridges is the key for enhancing the BLOS.

  • The BLOS of about 86% of the studied intersection is estimated as “C” or inferior.

Abstract

Introduction

Bicycling reduces the risk of various health problems associated with sedentary lifestyles. Hence, it is important to encourage bicycle commuting by enhancing the bikeability of transportation facilities. To support this process, this article proposes efficient bicycle level of service (BLOS) models for the assessment of signalized intersections under heterogeneous traffic conditions. Here, BLOS denotes bicyclists' perceived level of satisfaction.

Method

Extensive data sets (geometrical, traffic operational and built-environmental) are collected from 70 well-diversified intersection approaches of India. All approaches are also rated by 200 on-site bicyclists based on their perceived satisfaction levels on a Likert scale of 1–6 (excellent–worst). The attributes having significant influences on these ratings are then identified through Spearman's correlation analysis. Subsequently, three highly efficient techniques namely, associativity functional network (FN), genetic programming (GP) and step-wise regression are utilized to develop reliable BLOS models.

Results

As observed, the intersection BLOS is significantly (p < 0.001) influenced by total eight attributes of which crossing pedestrian volume, parking turn-over and average bicycle delay are the most dominating ones. Utilizing these variables, the FN tool has produced the most efficient BLOS model with a coefficient of determination (R2) value of 0.92 with averaged observations. Further, the classification of BLOS ratings into six symmetrical levels A–F (excellent–worst) has reported that around 86% intersection approaches in India are offering BLOS C–F.

Conclusion

The important measures of BLOS improvement at signalized intersections include the efficient management of crossing pedestrians, restrictions on nearby parking activities, and minimization of bicycle delay. The deficiencies in these aspects have perhaps made the intersection approaches in India to offer BLOS C–F. The BLOS models and transportation engineering solutions proposed in this study for the improvement of public health through bicycling are highly efficient for developing countries.

Introduction

Bicycling is a form of physical activity, which has a well-established connection with public health. It offers various health-related benefits such as increased cardiovascular fitness, improved joint mobility, strengthened bones and decreased body fat levels, etc. (Oja et al., 2011). Hence, the mobility requirements of bicyclists should be given due consideration in the infrastructure planning process to encourage bicycle use and improve public health. Transportation planners and engineers should actively work for enhancing the bikeability of existing facilities and establishing new bicycle-friendly road infrastructures. The bicycle service quality offered by roadways largely decides the number of bicycle trips generated from and attracted to different areas. Hence, the service quality estimation and necessary actions to improve the same are very important. However, to support these processes, suitable methodologies are yet to be developed for the contexts of developing countries. Hence, the present study has the following major objectives: (1) to identify the sets of attributes having significant influences on perceived service quality of through bicyclists at signalized intersections, and (2) to develop efficient bicycle level of service (BLOS) models for the prediction of bicyclists' perceived level of satisfaction. The results of these analyses are highly important, while the improvement of public health through bicycling is concerned. The BLOS models would assist the transportation planners and engineers in assessing the bikeability of signalized intersections, and recognizing the extent of needs for further improvements. These models could also be used to identify the inefficient and hazardous points on the road networks. Subsequently, the bikeability improvement strategies proposed in this study could be applied to make the whole road network bicycle-friendly. Such improvement works would largely encourage the inhabitants for bicycle use and improve their health conditions.

Previous studies (for instance, TRB, 2000; TRB, 2010; Jensen, 2013)) are mainly carried out under homogeneous traffic conditions prevailing in developed nations. The outcomes of these studies are not transferable to developing countries, where the road traffic symbolises a mixed flow of small and big vehicles with weak lane discipline, called heterogeneous traffic. In addition to such conditions, the unavailability of separate bicycle signals and bicycle-friendly facilities also makes the bicycle operation complex (Beura et al., 2017). This study analyses these conditions, and proposes efficient BLOS models along with various transportation engineering solutions for the improvement of public health in developing countries. For analysis purposes, a wide range of data is collected from various parts of India, and BLOS models are developed using genetic programming (GP) (Koza, 1992), associativity functional network (FN) (Castillo et al., 2000) and step-wise regression (Field, 2013) techniques. GP and FN are two highly efficient artificial intelligence (AI) techniques, which have numerous advantages over other prediction tools. These techniques develop predictive models based on the data knowledge without assuming or imposing any specific kind of relation between the independent and dependent variables beforehand, called data-driven approaches. Hence, these techniques are highly suitable for analyzing high dimensional problems with large data sets (Koza, 1992; Castillo et al., 2000). The prediction efficiencies of BLOS models developed using these techniques are evaluated using various statistical parameters, and the best model is reported. These models and other outcomes of this study are altogether an efficient support system, using which effective decisions can be taken at the planning stage itself for the betterment of bicyclists and public health. Complete detail on the study setting has been framed in Fig. 1, which mainly includes eight different steps as shown. A detail discussion on all steps are given in the following sections.

Section snippets

Background studies

Davis (1987) proposed the first-ever model for the bicycle safety evaluation at road intersections, called Intersection Evaluation Index (IEI) model. This model is a function of traffic volume, number of lanes, type of signalization, presence of turning lanes, curb radius and sight distance. The Highway Capacity Manual (HCM-2000) considered ‘controlled bicycle delay’ as the only measure of effectiveness (TRB, 2000). However, Crider et al. (2001) reported that the comfort levels of bicyclists

Brief descriptions of modelling tools

Of the three modelling tools used in this study, the step-wise regression analysis is a well-known statistical tool, whose analytical procedure could be found elsewhere. However, the working principles of GP and associativity FN techniques are not very common to the transportation engineering professionals. Hence, these techniques are briefly discussed below.

Site selection and data collection

The collection of diversified data sets (covering all possible conditions) is the key for developing a well-generalized model for any context. In this regard, 70 diversified intersection approaches are selected for investigation purposes. Here, the term “diversity” refers to the wide variability in the prevailing conditions of the intersections. To achieve this, study sites were selected from diversified locations (central, sub-urban and outskirts of cities), geometrical characteristics,

Variable selection, model development, and results

In the preliminary step, the variables having significant impacts on the perceived BLOS score were identified through Spearman's correlation analysis. This technique was opted as it can analyze both continuous and ordinal variables. In this analysis, various intersection attributes and bicyclists' characteristics (socio-demographic and travel-related) were used as independent variables, while perceived BLOS scores were used as the array of dependent variables. As bicyclists' characteristics are

Conclusions

Dealing with the BLOS analysis of signalized intersections, this study has arrived at various transportation engineering solutions for the encouragement of bicycling activities and improvement of public health in developing countries. The preliminary data analysis has revealed that the riding quality of through bicyclists at signalized intersections is primarily influenced by WEff, PHV, CPV, VTurn, D, PT and SDP. Thus, these variables are only used as input variables for the development of

Financial disclosure

The Authors did not receive any specific funding for this work.

CRediT authorship contribution statement

Sambit Kumar Beura: Conceptualization, Data curation, Formal analysis, Methodology, Software, Validation, Writing - original draft, Writing - review & editing. Kondamudi Vinod Kumar: Data curation, Writing - original draft. Shakti Suman: Formal analysis, Validation. Prasanta Kumar Bhuyan: Supervision.

References (22)

  • J. Strauss et al.

    Cyclist activity and injury risk analysis at signalized intersections: a Bayesian modelling approach

    Accid. Anal. Prev.

    (2013)
  • Aashto

    A Policy on Geometric Design for Streets and Highways

    (2011)
  • S.K. Beura et al.

    Operational analysis of signalized street segments using multi-gene genetic programming and functional network techniques

    Arabian J. Sci. Eng.

    (2018)
  • S.K. Beura et al.

    Level of service for bicycle through movement at signalized intersections operating under heterogeneous traffic flow conditions

    Transport. Develop. Econ.

    (2017)
  • D. Carter et al.

    Bicyclist intersection safety index. Transportation Research Record

    J. Transport. Res. Board

    (2007)
  • E. Castillo et al.

    A general framework for functional networks

    Networks

    (2000)
  • E. Castillo et al.

    Functional equations in applied sciences

    Math. Sci. Eng.

    (2004)
  • L.B. Crider et al.

    Multimodal LOS-point Level of Service Project, Final Report

    (2001)
  • J. Davis

    Bicycle safety evaluation

  • A. Field

    Discovering Statistics Using IBM SPSS Statistics

    (2013)
  • Guidelines for capacity of urban roads in plain areas

  • Cited by (11)

    • Electric bike (non)users’ health and comfort concerns pre and peri a world pandemic (COVID-19): A qualitative study

      2021, Journal of Transport and Health
      Citation Excerpt :

      In the cycling research domain, Davis (1987) documented the first so-called cycling LOS/safety index (Davis, 1987). This concept has since been further circulated in the field of cycling (Beura et al., 2020; Botma, 1995). The main process involved in LOS estimation is to aggregate/model the important variables that are associated with users' comfort and present them on a letter scale, often from A (best) to F (worst) (Jensen, 2007).

    • Quantifying changes in bicycle volumes using crowdsourced data

      2022, Environment and Planning B: Urban Analytics and City Science
    View all citing articles on Scopus
    View full text