SmartRoads – Automatic Accident Detection

Kin Fai Chan — Fri, 12 Jan 2024 10:40:53 +0000

Scenwise has carried out different projects for Rijkswaterstaat to show the potential of using traffic data to detect accidents and to speed up the incident management process. In 2019, we conducted different pilots for Rijkswaterstaat using loop data to detect accidents. This showed the potential of the algorithm. In 2021-2022, we conducted operational trial to demonstrate real-time validation of the Waze alerts. In more than 90% of the cases, our algorithm can correctly validate the waze alerts. This is valuable for speeding up the incident management process.

”An intelligent road which automatically warns the traffic manager and emergency services whenever an accident takes place ...

In 2023, in a follow-up project funded by the Eureka, Scenwise, Bureau Onderweg, ViNotion and Transnomis join force to develop an advance accident detection algorithm SRAIR (SmartRoads Automatic Incident Recognition). SRAIR combines a large variety of data sources including loops data, Floating Car Data (FCD), Waze alerts, NDW alerts, and video analytics data to achieve real-time accident detection on arterial/provincial roads.

Province of Noord-Brabant is highly interested in the potential of SmartRoads. An operational trial on the N279 was completed from September to December 2023. In 2024, a second pilot including multiple regions were successfully completed. We presented the results during the ITS Europe congress in Sevilla and different countries showed interest in our solutions.

Image source: Scenwise

Het bericht SmartRoads – Automatic Accident Detection verscheen eerst op Scenwise.

Traffic Jam Detection

Kin Fai Chan — Sun, 14 Feb 2021 06:33:01 +0000

Image source: unsplash

O ne of the tasks of Rijkswaterstaat (Dutch Highway Agency) is to provide actual and reliable traffic jams information to the road users. In the past, these traffic jams information was computed using the loop data.
This method has several limitations, e.g.

no nationwide coverage: there is no traffic jam information on the routes where there is no loop.
extra delays (> 3 min) due to the long information chain with multiple systems: MTM – Monica – Monibas – Trefi – TOS.
high maintenance costs because of the multiple systems in the information chain.
low resolution: heads and tails of the traffic jam can only be determined at the loop locations (per 500 m)

That is why Rijkswaterstaat and NDW wants to replace the outdated information chain with a new algorithm. Our consortium (Scenwise, Bureau Onderweg and Vortech) is the winner of this tender.

The new traffic jam detection algorithm is based on FCD. The new approach has the following advantages:

It has a nationwide coverage, even in areas where there is no lane signaling management (i.e. without loop detectors);
The length of the traffic jams and the heads and tails of the traffic jams can be determined with higher accuracy thanks to the high resolution of FCD (every 50 m).
The information chain is much shorter. This means more up-to-date data (1 min): The new solution replaces the processing steps of the current MTM-Monica-Monibas-Trefi information chain;
The solution is scalable to the arterial roads and urban road network. We also demonstrated this during the pilot.

The project was carried out with fixed-time and fixed-price agreement. Both Rijkswaterstaat and NDW are very satisfied with our scrum/agile development method and the final results. The algorithm was put into production shortly after the project was completed. The new algorithm has been in service since 2021.

Our detection algorithm is constantly improved using influences by government projects and subsidised projects. For example:

Eureka Investment programme pitch: Scenwise Vehicle Optimisation

GUIDE: GUIDE-pitch-deck

Het bericht Traffic Jam Detection verscheen eerst op Scenwise.

FCD Archieven • Scenwise

SmartRoads – Automatic Accident Detection

Traffic Jam Detection