Related Experiment Video
Updated: Mar 21, 2026

Evaluating the Effect of Roadside Parking on a Dual-Direction Urban Street
Published on: January 20, 2023
Speed-profile-based road segmentation for accident occurrence modelling for hilly terrains
Surya Prasanna Kumar Balusu1, Naveen Chandra Iraganaboina1, Regulus Dominic K Shallam1
1a Civil Engineering Department , Indian Institute of Technology , Guwahati , India.
A new method segments hilly roads using vehicle speed profiles, improving accident data accuracy. This speed-profile-based segmentation better captures road characteristics than constant length sections for accident modeling.
Area of Science:
- Transportation Engineering
- Road Safety Analysis
- Geospatial Data Analysis
Background:
- Accurate road segmentation is crucial for modeling accident occurrences per unit length.
- Traditional homogeneity-based and constant length segmentation methods are inadequate for hilly terrains due to geometric variations and ignoring road characteristics, leading to biased accident data.
- Vehicle speed profiles are hypothesized to reflect road geometry and roadside environment under free-flow conditions.
Purpose of the Study:
- To propose a novel methodology for segmenting road stretches in hilly terrains using vehicle speed profiles.
- To compare the effectiveness of speed-profile-based segmentation against constant length segmentation for accident data modeling.
Main Methods:
- A test vehicle (passenger car) was used to collect speed profile data on a hilly terrain road.
- A new segmentation methodology was developed based on analyzing the vehicle's speed profile.
- Statistical models were developed using both the proposed speed-profile-based segmentation and traditional constant length segmentation for comparative analysis.
Main Results:
- The speed-profile-based segmentation approach yielded statistically better results compared to the constant length segmentation approach.
- More parameters were found to be statistically significant in the model utilizing speed-profile-based segmentation.
- Road grade length, a key parameter in hilly terrains, was only statistically significant in the model based on speed-profile-based segmentation.
Conclusions:
- Speed-profile-based segmentation is a more effective method for segmenting hilly roads for accident analysis.
- This approach provides a more statistically robust foundation for accident modeling by better incorporating road characteristics.
- The methodology offers improved insights into the influence of road geometry, such as grade length, on accident occurrences.
Related Concept Videos
Design Example: Alignment of a Road Line Using GIS
Profile Leveling and Cross Sections
Elevation of Intermediate Points on Vertical Curves
Topographic Surveying and Contours
Design Example: Analyzing Capacity Contours for Flood Risk Assessment
Methods of Obtaining Topography

