Success Case:
Niedersachsen Ports
AI-POWERED RAIL TRACEABILITY
RaDaR4.0 gives us planning security, because we leverage comprehensive data through a single system made up of standard cameras and specialised software. — Holger Banik, CEO, Niedersachsen Ports
About Niedersachsen Ports
Niedersachsen Ports (NPorts) is the public port authority of the Lower Saxony region (Germany), headquartered in Oldenburg. With a total of 15 port locations, it is the largest operator of public seaports in Germany. Its portfolio includes five seaports, seven island supply ports, and three regional ports along the Niedersachsen coastline. Established as a GmbH & Co. KG, the company was founded in 2005 and is owned by the State of Lower Saxony.
The Challenge
Which wagon is on which track at the port of Brake? Until recently, answering that question meant sending NPorts operators on what amounted to a treasure hunt.
“From the rail companies we receive notification lists, but they are not always complete. Sometimes all we have is a blurry photo the driver sends from the road.” — Romina Hanisch, RaDaR4.0 Project Manager, NPorts
The consequence was direct: holding tracks in the port area were regularly over- or under-utilised whenever the actual number of wagons did not match the declared figure. In ports with a branching rail network and no central visibility — as is the case in Brake — the situation quickly became unmanageable. For operators, that meant additional work and, ultimately, a significant economic cost.
NPorts’ rail network across Brake, Cuxhaven, Emden, and Wilhelmshaven spans approximately 100 kilometres, with around 130 trains running through it each week. Managing that volume without reliable real-time data was a structural constraint on port efficiency.
The Solution
NPorts selected AllRead as the technology partner for the RaDaR4.0 project, funded by Germany’s Federal Ministry for Digitalisation and Transport (BMDV) and supervised by TÜV Rheinland.
AllRead’s proposal differed from the traditional OCR solutions available on the market — large portal capture structures similar to motorway toll gantries, expensive and poorly adaptable — by taking a lightweight approach: using surveillance cameras already installed at the ports, or fitting standard cameras several metres from the tracks.
As each train passes, AllRead’s computer vision software (OCR) automatically captures:
- Number of wagons and UIC codes
- Container numbers (BIC codes)
- Dangerous goods signage
- Direction of travel and transit time
Data is processed in real time and made available to operators, with image captures included for manual verification where needed.
“We deliberately chose this solution because it is more flexible for us.” — Romina Hanisch, NPorts
The Result
The project concluded successfully in June 2024. For operators, the impact is immediate and tangible. Before AllRead’s implementation, locating a specific wagon in the track network meant consulting incomplete lists, making phone calls, or even travelling physically to the port area. Today, more than 50 operators access AllRead’s platform to check in real time which trains have entered or left, how many wagons they comprise, which containers they are carrying, and whether any dangerous goods are on board — all of it backed by image captures for verification.
Planning work that was previously reactive and error-prone becomes a proactive process based on verified data. Less time spent searching for information, fewer track assignment errors, and a smoother logistics chain from the first wagon to the last.
One of the project’s differentiating factors was deployment speed. Traditional OCR solutions require the construction of heavy infrastructure with installation timelines that can stretch over months or even years. In the case of RaDaR4.0, the work was limited to running fibre optic and electrical cabling to certain remote points, installing poles with cameras and lighting, and connecting them to dedicated servers on the local network — no major civil works, no proprietary hardware. With that approach, AllRead deployed 18 control points across 5 ports in just over 6 months.
If you would like more information about this case, please contact our team.