Case studies

Analysing Flanders road safety with a traffic accidents map

Flanders is in Europe somewhere in the middle group regarding traffic safety. Traffic safety is a top priority for many citizens. A safe environment for pedestrians and cyclists is an absolute condition to increase the use of car alternatives in urban environments.

Reducing the number of accidents demands specific coordinated measures combining behaviour, infrastructure, enforcement and regulations. Information about the specific location of traffic accidents is an essential element to reach the Flemish goals on traffic safety.

The draft Flanders Mobility Plan (2015) defines five clear goals for 2030 (yearly):

Maximum 133 deadly injured persons;
Maximum of 1000 severely wounded persons;
Maximum of 540 fatal and severely injured persons in the group of non protected road users (pedestrians, cyclists, motorcyclists);
Maximum 350 deadly and severely injured car drivers in the group of road users between 18 and 24 years old and the group of 65+ years old;
Reducing the number of people with light injuries with at least 20%.

The Flemish traffic safety programme focus on five pillars. Each pillar is linked to policy actions.

Pillar 1: Informing, sensitising and training for safer traffic behaviour;
Pillar 2: traffic-safe design and use of traffic-safe technologies;
Pillar 3: enforcement and regulations ensure a sufficiently effective road safety policy;
Pillar 4: evaluation and monitoring to ensure an effective safety policy;
Pillar 5: the need for engagement: behavioural change.

This initiative contributes to Pillar 2 (Action improve the safety of dangerous places and road segments and blind corners), Pillar 3 (ANPR and trajectory speed control) and Pillar 4 (Policy-relevant research, evaluation and monitoring).

Besides the direct impact, the overview map also makes people aware of the traffic safety problem. The map makes it very easy to select several traffic safety-related issues as accidents during the weekend nights, during the morning and evening rush hours from Monday to Friday when young children going to school, blind corner accidents or accidents where alcohol and drugs are involved.

Challenges encountered:

Until 2016 the traffic accident data (even it was anonymised) was considered as data for internal use. There was no access to individual anonymised accident data for non-professionals. The main concern for non-opening was that the data itself – when shown on a (heat)map – was considered ‘sensitive’ in a way that the information could lead to misinterpretation. This concern has been solved by combining the heatmap functionality with the exact location of the accidents and by adding an introductory text providing a view on the amount of missing (location)data.

A second challenge was improving data quality. Since the data is an interpretation of the on terrain information, in a lot of cases crucial accident information is not well inputted on the terrain, and the geographical location needs to be interpreted too because no GPS coördinates are recorded. The data quality and especially the exact location was derived from post addresses, crossings and km points along the main roads. All of the points were located on the road itself.

A third challenge was to combine the accident data with other interesting datasets to make it more relevant without losing focus on road safety and traffic accidents. By selecting only the most relevant datasets, the story stays crisp and clear. The selection of what must be shown in what way can be best made together with the stakeholders keeping in mind the less is more approach.

An important element in convincing the Federal Police to visualize the data in a public interface was the existence of a traffic accident map example build in the H2020 Open Transport Net EU project for the city of Birmingham (UK).

Stakeholders:

Informatie Vlaanderen: Responsible for the data visualisation;
Federal policy: Responsible for data integration and data delivery;
Local policy districts: Responsible for the data acquisition on the terrain;
Mobility and public works department Flanders: Responsible for the traffic safety policy and planning;
Road and traffic department Flanders: Responsible for the Flemish road infrastructure including trajectory control zones;
Flemish traffic safety foundation: Responsible for traffic safety actions;
VIAS institute.

Actions steps:

Negotiating with the Federal police to set up cooperation;
Functional analysis to build the traffic safety and accident map;
Adding new datasets that support the policy problem: School locations.

Future Actions:

Adding trajectory control locations;
Adding speed camera locations;
Adding predefined scenarios (weekend accidents, accidents during rush hours, blind spot accidents, trajectory control zones effects,…);
Testing the usability of the map & data literacy;
Integrate the map in (local) mobility dashboards.

Lessons learned:

Demonstrating the possibilities of a dataset helps to open the datasets – showing other traffic safety maps helps to convince data providers that it is important to open their data;
Limit the number of datasets in one map to keep the policy story clear;
One single map can contribute to several policy actions;
Pre-defined scenarios help to make an advanced map easier to consume.

Outcome impact:

The traffic accident map is used as one of the instruments to define the list of dangerous places and road segments;
The map is also used as one of the instruments to decide about the priority and place where new trajectory controls will be installed along the Flemish main roads.