Ports are essential to the global economy, serving as key hubs for international trade. Port efficiency is crucial for the smooth flow of supply chains.

However, the increase in port traffic presents significant challenges, particularly in mitigating the environmental impact of docked ships and vehicles within terminals. Emission reduction is a priority for many countries, with the International Maritime Organization (IMO) aiming for zero emissions by 2050.

Solutions include optimizing logistics operations and technological innovation in ports. A key indicator of port efficiency is dwell time, which measures how long a container stays in the port before being transferred.

A high dwell time indicates logistical inefficiencies and a negative environmental impact, with increased emissions and noise pollution caused by congestion.

Poor communication between port operators and transport companies contributes to this problem, causing disorganization, operational delays, and accidents. The proposed solution introduces a geolocation and visual identification system, using digital cameras and a convolutional neural network to detect the position of trailers within the port, recording cargo information and trailer license plates.

This data is managed through a web application that provides guidance on optimal routes, considering factors such as congestion and ongoing operations. The system, based on a reinforcement learning neural network, improves over time through feedback, adapting to the port’s dynamic changes.

This technology helps reduce emissions and enhance safety, promoting port logistics that are more efficient, sustainable, and secure.