Research

Research Areas 

• Road Traffic Safety 

• Driver Behavioral Study

• Crash Data Analysis

• Surrogate Safety Measures

• Traffic Flow Microsimulation

• Safety Performance Functions

• Traffic Conflict Technique

• Roundabout Safety

Research Work

1. Ph.D. Research Project

Title: Driving behavior at roundabouts from a safety perspective

Summary: Although roundabouts are considered considerably safer than other intersections, many accidents are still reported. The issue requires intensive research to understand the basis of such many accidents. Research is lacking due to a shortage of crash data for analysis. The research methodology includes deploying drones for video capture, employing unsupervised learning for traffic conflict detection, identifying causal factors of roundabout crashes, developing a microsimulation model for safety assessment, and creating a conflict-based safety performance function. The ultimate goal is to recommend countermeasures for reducing crashes at roundabouts. Using surrogate safety measures such as traffic conflict analysis, the underlying issues must be highlighted. The research addresses existing research gaps and emphasizes the importance of implementing site-specific roundabout safety solutions.

2. M.Tech Research Project

Title: Service level modeling of motorized three-wheelers at un-signalized intersections under heterogeneous traffic flow conditions

Summary: The research primarily focused on developing Three-Wheeler Level of Service (3WLOS) models to assess service quality at uncontrolled un-signalized intersections under heterogeneous traffic conditions in developing countries like India. Data from 21 intersections in seven cities across five states in India were analyzed to identify significant variables affecting driver satisfaction. Nine key factors, including pavement condition, roadway width, median barriers, land use, and others, influenced 3WLOS scores. Various statistical and artificial intelligence techniques were employed to predict 3WLOS scores, with Bayesian Regularized Artificial Neural Network (BRANN) performing the best. Sensitivity analysis revealed the service delay had the highest negative impact on driver comfort, highlighting the need to address delays at uncontrolled intersections for improved driving conditions.