Sure! Here’s the translation into American English:
—
A team from the Institute of Information and Communication Technologies (ITACA) at the Polytechnic University of Valencia (UPV) and the Instituto de Física Corpuscular (IFIC) has presented an innovative prediction and early warning system for urban traffic that utilizes deep learning techniques. This advancement aims to anticipate episodes of high pollution, thereby facilitating the implementation of preventive measures to improve air quality in cities.
Traffic-related pollution accounts for approximately 60% of greenhouse gas emissions in cities like Valencia. With the new system, it is possible to forecast 30 minutes in advance whether a street segment will reach elevated traffic levels, enabling authorities to make quick decisions to mitigate pollution and safeguard public health.
Edgar Lorenzo-Sáez, a researcher at ITACA, emphasizes the seriousness of air pollution, which is the leading environmental cause of premature deaths. This phenomenon has been linked to serious health issues such as asthma and cardiovascular diseases, causing approximately 300,000 deaths annually in the European Union.
The system has been trained with data obtained from 1,472 traffic sensors in Valencia, and it also includes meteorological variables. This model classifies each road segment into three alert levels, demonstrating impressive accuracy in real-time thanks to LSTM neural networks, even during peak hours. Furthermore, traffic data has been shown to be a reliable indicator of nitrogen oxides (NOx) levels, one of the most harmful pollutants for health.
The system’s effectiveness could be crucial in improving Low Emission Zones, allowing decisions to be made in a more localized manner, tailored to the actual risk of each area. Lorenzo-Sáez highlights that the system is correct 90% of the time under flowing traffic conditions and 70% when anticipating episodes of heavy traffic, opening up new possibilities for more flexible pollution management.
Javier Urchueguía, also from ITACA, mentions the direct relationship between traffic flows and NOx levels, which is an essential factor for cities with limited resources that do not have a complete air quality sensor network. Verónica Sanz, a professor at UV and co-author of the study, adds that artificial intelligence models can adapt to different scenarios, making them applicable in various locations.
This development represents a step toward more effective and sustainable urban management, integrating artificial intelligence as a key tool to tackle complex environmental challenges. In the future, there are plans to create a digital twin of the city of Valencia, allowing for the simulation of measures before implementation and increasing the number of Internet of Things sensors to enhance pollutant prediction. The goal is not only to contribute to the sustainability of cities but also to promote the health and well-being of their inhabitants.
—
Let me know if you need any further changes!
via: MiMub in Spanish
