Dutch rail infrastructure manager ProRail believes that intelligent cameras, in combination with wheel passage sensors, can generate up to 25% more capacity at freight yards. It has issued a tender for cameras that can automatically recognise wagon numbers and hazardous materials, among other things.
The capacity problem primarily affects marshalling yards that do not (sufficiently) know the actual occupancy and utilisation of the tracks, according to ProRail. Previous pilots have shown that a combination of wireless wheel passage sensors and intelligent cameras can improve this. ProRail now intends to apply this technology on a larger scale in the Netherlands.
The first implementations will take place at three freight yards: Houtrakpolder, Moerdijk, and Maasvlakte West/Westwest. Three, one, and five camera systems will be installed there, respectively. In addition, ProRail is considering an expansion to Barendrecht Vork/Vaanplein and Sloe/Vlissingen. Other passenger and freight yards may also be brought under the same agreement at a later date.
The goal is not only to know where trains and wagons are located. The data is intended to help ProRail use available parking capacity more efficiently and to better register hazardous materials. The rail infrastructure manager expects that this will ultimately grow the plannable capacity for goods by 25%, without laying extra tracks.
Complete assignment
For the supplier, the tender means more than just installing cameras. The contract also includes installation, management, maintenance, troubleshooting, and the analysis and delivery of the collected data. Moreover, the systems must be able to function in an operational railway environment.
In addition, ProRail is explicitly seeking a solution that can be further developed. For instance, the railway infrastructure manager is asking about the possibility of automatically sending camera images to the ProRail cloud environment in the future. Bidders must also indicate how future developments can be incorporated into a roadmap.
The technical requirements are stringent. Bidders must have recent experience with smart cameras and intelligent video analysis in a rail environment or similar critical infrastructure. Reference projects must, among other things, have demonstrated an object detection accuracy of at least 95%.
Price accounts for 30% of the assessment; quality accounts for 70%. A ceiling price of €1 million excluding VAT applies to the first lot.
Following the provisional award, the prospective winner must additionally test the system for three to four weeks at the Moerdijk site. The final award will only follow after a successful quality test. The agreement is scheduled to take effect on 1 February 2027 and has a term of five years, with three options for a one-year extension.
This article was originally published by our sister publication SpoorPro.nl.