Success Case:
Hutchison Ports Best
RAILWAY ACCESS MANAGEMNET
It was a project with a high level of collaboration between the teams, which allowed BEST to continue advancing with the digitalization and optimization of its processes. The data is integrated in a simple way, and the logistics team receives it almost in real time. - Santiago Pallares, Operatios Development Manager in BEST
About Hutchison Ports BEST
Hutchison Ports BEST is the first semi-automated terminal of the Hutchison Ports Group and the most technologically advanced port development project in Spain. With a handling of 2,500,000 TEUs per year, BEST is a reference for terminals around the world.
The Challenge
Best manages around 3,000 entries and exists of trucks and 20 trains per day at its facility. Despite being a leader in technological advances, BEST encountered a major obstacle in terms of the efficiency of monitoring freight trains and loading. With around 15 per cent of the traffic being carried by trains, the manual process of verifying and capturing essential data during loading and unloading was taking as long as 40 minutes per cycle. This slow process impeded operational efficiency and competitiveness, creating bottlenecks in the terminal’s workflow. To improve its performance, without impacting the terminal’s daily operations, BEST needed a solution that:
- Extract the Container BIC Codes.
- Extract the UIC Wagons.
- Give as an output the complete “Train Map”.
- Is easily integrated with existent systems.
The Solution: Rail ARS Software
Understanding the critical need for a more efficient and error-free data capture system, BEST collaborated with AllRead on the setup of its system for automatic recognition of ins and outs of containers by rail. As such, AllRead provided its software (Agile Recognition Software – ARS) with On Premise integration and low hardware needs, allowing the verification and data capture process to be streamlined.
With the installation of two IP cameras with specialized lighting at the terminal entrance, AllRead ARS began to extract vital data for the operation, integrating it directly into the BEST system. This seamless integration ensured continuous loading and unloading operations without creating operational bottlenecks.
The Result
The implementation of AllRead ARS 2.9, powered by deep learning and computer vision, brought remarkable improvements by rapidly processing the images, providing complete data on train composition. From 40 minutes of data capture processing time to 10 minutes of data passage, an 80% improvement in efficiency, enabled BEST to manage train arrivals and departures both day and night more effectively.
The project produced exceptional results: automatic wagon readings reached 95%, container readings 96% and train association and mapping a staggering 98%. These results underline the robustness of the software and its reliability for application in critical BEST processes.