Pre-crash scenarios at road junctions: A clustering method for car crash data

Abstract

Given the recent advancements in autonomous driving functions, one of the main challenges is safe and efficient operation in complex traffic situations such as road junctions. There is a need for comprehensive testing, either in virtual simulation environments or on real-world test tracks. This paper presents a novel data analysis method including the preparation, analysis and visualization of car crash data, to identify the critical pre-crash scenarios at T- and four-legged junctions as a basis for testing the safety of automated driving systems. The presented method employs k-medoids to cluster historical junction crash data into distinct partitions and then applies the association rules algorithm to each cluster to specify the driving scenarios in more detail. The dataset used consists of 1056 junction crashes in the UK, which were exported from the in-depth "On-the-Spot" database. The study resulted in thirteen crash clusters for T-junctions, and six crash clusters for crossroads. Association rules revealed common crash characteristics, which were the basis for the scenario descriptions. The results support existing findings on road junction accidents and provide benchmark situations for safety performance tests in order to reduce the possible number parameter combinations.

Identifying Seriousness of Car Accidents using Data Mining in Jordanian Roads

“…Derived from one of the busiest roads within Nigeria, Olutayo & Eludire [9] examined Nigerian traffic accidents by organising the obtained data into continuous-examined using the Artifical Neural Network-and categorical-examined using the decision tree technique-data. Implemented on three-and four-legged UK road junctions, Nitsche et al [10] additionally suggested a new strategy on taking essential pre-crash situations from accident data. Considering it is capable of handling categorical data and is strong against outliers, the clustering method k-medoids was discovered to be the best for the dataset at hand [11].…”

A novel, modular validation framework for collision avoidance of automated vehicles at road junctions

“…This work provides three novel contributions to existing approaches. First, the junction scenarios to be evaluated are obtained from accident data by using a combination of clustering and association rule mining, as published by Nitsche et al [21]. A database-driven method to follow this holistic approach was developed by Pütz et al [12].…”

“…This section briefly describes the accident data mining approach. For a more detailed explanation of crash data attributes and results of the method, refer to [21], where this 2-step approach was applied to an in-depth junction accident dataset. Basically, the aim of the crash data analysis and mining is to identify distinct groups of safety-critical scenarios relevant for the test.…”

Methodology Designed to Evaluate Accidents at Intersection Crossings with Respect to Forensic Purposes and Transport Sustainability

“…An important consequence of transport sustainability will be improved traffic safety [1]. Given the recent advancements in autonomous driving functions, one of the main challenges is the safe and efficient operation of complex traffic situations such as road intersections [2]. Over the past few years, the automation of road vehicles has gained an increasing presence on the agendas of companies and public authorities, which have started to push for automated driving systems (ADS) to the forefront of research [2,3].…”