Proactive Pedestrian Safety Evaluation at Unsignalized Intersections in India Using Surrogate Safety Measures

The safety evaluation using reactive approach (conventional method) requires crash data of several years. However, in developing countries like India, the reliability and accuracy of the available accident data are highly questionable. Hence more effective and proactive safety evaluation technique is required. Therefore, the objective of this work is to propose proactive surrogate safety measures based methodology to quantify pedestrian safety at the unsignalized intersections in India. The main advantage of the proximal safety indicator is the incorporation of the conflicts more frequently than the actual accidents, which results in the effective, statistically significant and reliable proximal measure of traffic safety. The Post Encroachment Time (PET) is used as a proximal safety indicator in this study. The required pedestrian-vehicle conflicts were extracted manually from the videographic survey conducted at an unsignalized intersection. The conflicts were grouped as highly severe conflict, severe conflict, and normal conflict according to the behavior of the participants of the conflicts (pedestrian & vehicle). For each type of the vehicle class, the threshold PET values of each conflict group (highly severe conflict, severe conflict, and normal conflict) were obtained using Support Vector Machine algorithm. This paper describes the effect of vehicle type and the vehicle approaching speed on the severity of the conflict under non-lane based mix traffic condition. Additionally, it also sets the threshold PET values for highly severe conflict, severe conflict, and normal conflict which can be used for the prediction of the severity of the conflict at unsignalized intersections in India.

• Supplemental Notes:
  • This paper was sponsored by TRB committee ABE90 Standing Committee on Transportation in the Developing Countries.
• Authors:
  • Shah, Harsh
  • Vedagiri, P
• Conference:
  • Transportation Research Board 97th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2018-1-7 to 2018-1-11
• Date: 2018

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 8p

Subject/Index Terms

• TRT Terms: Developing countries; Machine learning; Methodology; Pedestrian safety; Safety; Traffic conflicts; Unsignalized intersections; Vehicle mix
• Uncontrolled Terms: Conflict severity; Surrogate safety measures
• Geographic Terms: India
• Subject Areas: Highways; Operations and Traffic Management; Pedestrians and Bicyclists; Planning and Forecasting; Safety and Human Factors;

Filing Info

• Accession Number: 01656843
• Record Type: Publication
• Report/Paper Numbers: 18-02052
• Files: TRIS, TRB, ATRI
• Created Date: Jan 22 2018 10:49AM