The Port of Barcelona and the startup AllRead use artificial intelligence to identify containers and wagons

AllRead Machine Learning Technologies (MLT), a startup specialized in computer vision software (Computer Vision), has successfully developed a pilot project at the Port of Barcelona to identify containers and wagons with its artificial intelligence-based technology.

The company created in The Collider, the innovation program of Mobile World Capital Barcelona, uses technology based on artificial intelligence (neural networks) to process images and videos with text and codes, and extract only the desired information. The technology has been developed at the Computer Vision Center of the Autonomous University of Barcelona (UAB).

The collaboration arose from a challenge launched by the Port of Barcelona to identify containers and wagons entering the port area by applying AllRead's reading technology. The objective was to achieve maximum reading accuracy, without having to install large additional infrastructures, such as the usual steel portals with multiple cameras and hardware equipment. The results of the project have been very positive, achieving an accuracy level of up to 99% of the identified assets.

The project, which started at the beginning of 2020 and ended in February, demonstrated by analyzing video files that it is possible to return the container and wagon codes, either vertically, horizontally, or combined, and even when equipment was dirty, broken or partially covered. With further processing, it was also possible to create a mapping of the train, creating relationships between wagons and containers. The tests have been carried out both during the day and at night, to check the operation of the system in low light conditions.

In the words of Miguel Silva-Constenla, CEO of AllRead, "with this technology, we seek to promote the ubiquity of traceability in port facilities, removing financial and technological barriers to its adoption."

The AllRead MLT solution, installed in two chambers, has analyzed a total of 950 trains and 13,500 containers, providing the Port of Barcelona - a pioneer in the adoption of this technology in Spain - with automatic control of one of the rail lines at the Hutchison Ports BEST terminal. The technology contributes, according to the Port's objectives, to optimize operations and automate data collection of goods, favoring the processes of loading and unloading trains in the port area.

The implementation of AllRead MLT technology is part of the commitment to innovation and digitization that the Port of Barcelona has been carrying out in recent years. Its president, Mercè Conesa, recalls that "innovation and digital transformation are our tools to move towards a sustainable future. They are two levers on which we rely to be an increasingly smart port. We want to be an agent of change on the road to smart and sustainable mobility and there is no doubt that artificial intelligence and technologies such as AllRead helps us achieve our goals."