A novel rare event approach to measure the randomness and concentration of road accidents

Abstract

BackgroundRoad accidents are one of the main causes of death around the world and yet, from a time-space perspective, they are a rare event. To help us prevent accidents, a metric to determine the level of concentration of road accidents in a city could aid us to determine whether most of the accidents are constrained in a small number of places (hence, the environment plays a leading role) or whether accidents are dispersed over a city as a whole (hence, the driver has the biggest influence).MethodsHere, we apply a new metric, the Rare Event Concentration Coefficient (RECC), to measure the concentration of road accidents based on a mixture model applied to the counts of road accidents over a discretised space. A test application of a tessellation of the space and mixture model is shown using two types of road accident data: an urban environment recorded in London between 2005 and 2014 and a motorway environment recorded in Mexico between 2015 and 2016.FindingsIn terms of their concentration, about 5% of the road junctions are the site of 50% of the accidents while around 80% of the road junctions expect close to zero accidents. Accidents which occur in regions with a high accident rate can be considered to have a strong component related to the environment and therefore changes, such as a road intervention or a change in the speed limit, might be introduced and their impact measured by changes to the RECC metric. This new procedure helps us identify regions with a high accident rate and determine whether the observed number of road accidents at a road junction has decreased over time and hence track structural changes in the road accident settings.

The heartbeat of the city

“…On streets with large amounts of traffic (both higher speed and traffic volume), more accidents are expected [ 27 , 73 ]. Therefore, since traffic is concentrated in certain areas, and since environmental factors and some human errors occur more frequently in some areas than others, crashes are concentrated and form non-random hotspot patterns [ 40 ]. These include, for example, areas with a high density of pedestrians in the dark, or where there are “careless weekend drivers” [ 39 ].…”

“…Using existing legal or political definitions for neighbourhoods or districts often yields a more natural partition of a city. Yet, the dividing line between two distinct neighbourhoods is often a main avenue, and the edge between three of more neighbourhoods is often a main city node (such as a Metro station) which functions as an attractor node of crime and crashes [ 40 ]. This results in crucial gravity centres of the city being divided into distinct observation units.…”

“…The fact that crime is concentrated results in hotspots of crime, which has many implications in terms of security [ 38 ]. Road accidents are also concentrated in space, and form hotspots [ 25 , 39 ], usually nearby highways and major junctions, that are not just the result of randomness [ 40 ]. The accident hotspots also give relevant patterns in terms of where they happen, for example, that most pedestrian injuries are situated on major roads [ 41 ].…”

Urban population size and road traffic collisions in Europe

“…Since Smeed's 1949 pioneering work regarding statistical aspects of road traffic collisions [34], the precision and availability of both geographic and road safety data have improved considerably, enabling many other authors to expand the field [35][36][37][38][39][40][41][42]. Additionally, the more recent introduction of scaling models in the context of urban science [43] offers a new avenue for modelling road traffic collisions and understanding their behaviour.…”

Predictors and geographic analysis of road traffic accidents in Leon, Nicaragua

“…14 The presence of intersections, trees that obstruct vision while driving or walking, and pedestrian space occupied by informal businesses can increase the risk of crashes. 15,16,25 In this study, a greater number of accidents occurred on asphalt roads that were in good conditions and on roads that contained bicycle lanes. This is in line with previous studies that showed association of adverse infrastructure conditions, such as road surface conditions, street type and road width, lighting as well as location type with higher odds to RTA.…”