AllRead launches the “Disruption of Port OCR” project financed by the Ports 4.0 fund of Puertos del Estado.
Optimised freight traffic monitoring
The artificial intelligence startup AllRead, in The Collider portfolio, has launched the “Disruption of Port OCR” project financed by Ports 4.0, the capital fund of Puertos del Estado and the Spanish Port Authorities. The goal of this 18-month project is to democratise the use of disruptive OCR in port facilities and overcome what poses one of the greatest challenges for many Spanish ports: automated traceability of rail traffic wagons and goods.
Deep learning and computer vision to democratise port OCR
OCR (optical character recognition) systems are essential to traceability in freight transport and logistics. They can read and digitize images, numbers, codes, etc., and make it possible to identify and control stock and monitor the flow of vehicles and containers. However, conventional OCR lack high precision and require a significant investment in hardware installation, maintenance and connectivity. For these reasons, they are not widely used by port facilities in Spain, especially at rail accesses.
AllRead has overcome this hurdle with its software based on deep learning and computer vision. The software detects and reads the texts or codes on number plates, trailers, containers, wagons or hazardous goods in real time, using any of the port’s existing cameras. This solution boasts a reliability of 99% thanks to its convolutional neural network architecture, the product of more than 5 years of research at the Computer Vision Centre (UAB). As a result, it works even in complicated conditions such as rain, poor lighting, low visibility or high-speed movement.
The software is lighter and easier to install than conventional OCR and reduces implementation costs by 80%. It is simple to maintain and offers a business model that adapts to the traffic volume of each port.
Efficiency and savings
The software improves operational efficiency and reduces bottlenecks, regardless of the size or volume of port operations. This solution also reduces delays and transit times by up to 25% and translates into savings of as much as €80M per year.
In addition to expediting the entry, exit and circulation of vehicles and goods, AllRead software contributes to sustainability by reducing CO2 emissions by 220,000 tonnes and promoting rail transport, which is six times less pollutant than road transport.