The police have changed their monitoring tactics.Continue reading
The Center for Budapest Transport (BKK) presented its “Vision Zero” project that aims to reduce the number of fatal traffic accidents by 50% by 2030, and to zero by 2050. To achieve this, BKK is also using its own AI-based software that can predict with an accuracy of over 90% how many accidents will happen on which sections of road, Index reports.
Ádám Bodor, Director of Mobility Development at BKK, explained that the goal set by Mayor Gergely Karácsony is to reduce the number of fatal and serious accidents to zero. The most important measures of the road safety strategy relate to the reduction of vehicle speed. One means of achieving this objective in a number of stages is speed control in the city. To do this end, the plan sets out the speed limits required by 2030, on major road sections in the capital.
BKK has classified the capital’s roads in detail according to the public function of the road section:
The larger a road is, the more vulnerable road users, i.e. pedestrians and cyclists, use it. Consequently, most accidents occur on road sections where both car and pedestrian traffic is heavy.
Using their artificial intelligence model, BKK’s data analysts have identified the road sections defined in the strategy where serious and fatal accidents can be avoided simply by reducing the speed limit. On this basis, the maximum speed on certain sections will be reduced from 50 km/h to 40 km/h. “A pedestrian is four times more likely to survive if they collide with a vehicle traveling at 30 km/h rather than 50 km/h,” said Ádám Bodor.
What this will mean in figures is shown by examples from abroad. BKK has adapted London’s example for Budapest, as the system is already in use in the UK capital. In London, the speed limits have reduced the number of road deaths by 46% and the number of serious and minor injuries by 25%.
The first street in Budapest to benefit from the speed reduction was a section of Attila Street, where the number of accidents has statistically fallen by more than half.
On some other busy stretches of road, the measure will be introduced this autumn, reducing and standardizing speed limits.
Photo: MTI/Balogh Zoltán
The work of BKK experts is also supported by the AI-based system they have developed. Patrik Tóth, head of the Modelling and Data Analysis Department at BKK, explained that they initially created an accident scene map based on police incidents, but this had to be done largely manually, therefore they thought about automating the process.
As a first step, they created a graph system that covered all road sections in the capital and was parameterized with their own and publicly available data. This data also included accident reports, which are made in the Waze app used by many people. The program was then trained using 17 criteria to create a predictive model that could be used to predict which sections of road to tackle first. The system was tested by calculating how many accidents will occur on which road sections in 2024, based on data from 2019 to 2023.
The results showed that the system could predict accidents with an accuracy of more than 90%.
The system was also used to determine which road sections would make the biggest difference if car speeds were reduced and will be used as the basis for the phased introduction of road changes for the project, which will run until 2030.
Via Index, Featured image: Pixabay
