Use our Rail OCR solution to automate the registration of train data. Increase the efficiency of your tracks, improve safety, and save on labor costs. Real-time video images are analyzed using IP cameras in combination with artificial intelligence. Our algorithm is able to quickly understand and recognize everything on the video.
Think of information for container terminals such as:
For bulk terminals, we recognize information such as:
Our Rail OCR solution is incredibly accurate because our technology works with video images. By using video instead of photos, the algorithm has the choice of thousands of individual frames of an object to base conclusions on. This means that the quality and reliability of the collected data with our technology is unmatched by traditional Rail OCR systems.
What makes our Rail OCR solution unique is that supplai does not require any detection system in the tracks. Our algorithms recognize the direction of the train, count the number of wagons, and can therefore handle shunting. Does a train stop? Does it decrease or increase speed? No problem!
Line-scan cameras are used for traditional Rail OCR, which take multiple photos at one height. These photos are then stitched together into one image and scanned based on a predefined set of letters and characters. Therefore, traditional Rail OCR systems can have difficulty reading information that is not placed in a standard location, or when the photo is unclear due to lighting or weather conditions. That is why traditional Rail OCR systems remain dependent on human controls to correct errors and ensure accuracy. Hardware also remains a major investment.
Artificial intelligence is now also used to analyze the photos taken with line-scan cameras. This still does not come close to the speed and accuracy of our technology.
Our Rail OCR solution uses artificial intelligence in combination with IP cameras, which makes it possible to read information from a real-time video. Our technology uses algorithms to understand and analyze the videos in a way that is close to human intelligence. The algorithm is self-learning and can adapt to become more accurate as it processes more information. Our Rail OCR technology is therefore very flexible and can handle different situations; think of different types of characters, trains, layouts, and languages, but also bad weather or busy backgrounds. Because our Rail OCR technology is software-based, the investment is also much lower!
ProRail even saw an opportunity to go one step further and, together with supplai, developed the Brake Shoe Model. An algorithm to measure the thickness of the brake pads. All based on AI and video without any sensors in the track. Read more about the collaboration between ProRail and supplai here.
An online platform is part of our Rail OCR solution. Here, all collected images and data are stored, and you can train the algorithm. For example, if the algorithm has not recognized something correctly, you can correct it in the platform. This way, you train the algorithm to understand it correctly the next time. This makes the quality of your Rail OCR solution better over time! Nice to have such a new smart employee, right?
The platform can be fully customized based on your preferences. For example, let it communicate intelligently with every TOS or TMS you use on your railway terminal. This way, you make your processes even more efficient! If damage is reported to a wagon, tank, or container, simply enter the relevant number and immediately see the correct video and photos!
The Rail OCR platform is even available for mobile devices and works on various operating systems.
Our algorithms are local to the terminal, which means that we do not need to send data to external servers to process it. This makes our solution faster and safer. Moreover, we meet all required security standards to ensure that our customers’ data is safe and protected!
Many terminals are under development and excavation work can be costly. That’s why we’ve innovated by creating a fully mobile portal. We achieve this by casting the portal in a block of concrete, making it easily movable. The portal we installed at Broekman Logistics, for example, is 18 meters long and only weighs 300 kg. It’s ideal if you anticipate changes to your terminal but still want to automate!
“We chose to collaborate with supplai because of their expertise and experience with Rail OCR, the speed at which the solution could be implemented, and the minimal changes required in our hardware.”