The Port of Barcelona has begun a pilot test of 5G Maritime, which aims to validate the location of ships in real-time through artificial intelligence (AI), cloud computing and edge computing.
The port pilot will be able to accurately estimate valuable information for the entry and berthing operations of vessels in the port, such as the location of several ships sailing within the port’s docks, the course or even the speed of the ship itself, all based on an analysis of video from 5G cameras installed around the port area.
Mercè Conesa, president of the Port of Barcelona, said: “Having a network of cameras connected by 5G technology would represent an important advance in terms of safety for maritime traffic in the Port of Barcelona.
“But not only would it help us improve safety and security in the port area, which are of the utmost importance, but it would also help to optimise our dockside management. And it would facilitate the daily work of all port services — pilots, tug-operators and moorers.”
As 9,000 vessels visit the Port of Barcelona annually, with some as long as 400 m, having precise real-time information on their movements and geolocation is essential to optimise docking space and maintain safety during manoeuvres according to the port authority.
The project is part of the 5G Barcelona initiative and promoted by the Port of Barcelona, IBM, Vodafone, Huawei, Mobile World Capital Barcelona and Fundació i2CAT, with the support of the Catalan Regional Ministry of Digital Policies within Catalonia’s 5G Strategy.
In particular, the solution features three elements including an AI model – IBM Maximo Visual Insights, trained on the IBM Cloud to recognise ships and their bow and stern.
It also has two sets of high-performance cameras and 5G terminals from Huawei located in the port, which capture the entry of ships in real time.
Finally, it uses ultra-fast, low-latency communications provided by Vodafone’s commercial 5G network that transmit the images to the server, where the images received are run through an AI model which recognises and interprets them.
Once the AI model is distributed to a server in the port control tower using IBM edge computing technologies, a geolocation algorithm translates the image pixels into geographical latitude and longitude coordinates.
This data is integrated with the rest of the port’s systems, thereby delivering exact information on the location of vessels in real-time.