AI-driven system uses real-time weather data to reduce harmful coastal pollution through fuel efficient ship navigation
BUSAN, South Korea, Aug. 14, 2026 /PRNewswire/ — Air pollution from ships poses a serious health risk in major port cities, yet existing mitigation strategies often rely on reducing the speed of sailing, which affects operational efficiency. Researchers from Pusan National University have now developed an AI-powered framework that combines real-time weather data, physics-informed learning and multi-objective optimization. It recommends optimal ship routes and speed profiles to reduce peak pollution exposure by predicting how exhaust plumes disperse, while maintaining fuel-efficient operations.
While maritime transportation forms the backbone of global trade, it is also a major contributor to air pollution in coastal cities. Although regulations such as cleaner marine fuels and speed restrictions have helped reduce emissions, nearby communities continue to face health risks because pollution levels depend not only on how much ships emit but also on how wind and weather disperse exhaust plumes. Conventional mitigation strategies often rely on blanket speed reductions, but these often increase travel time while failing to prevent short-term pollution spikes in densely populated areas.
To address this challenge, researchers from Pusan National University developed a physics-informed artificial intelligence framework that optimizes ship routes and speed profiles to simultaneously reduce fuel consumption and peak air-pollutant exposure during ship operations. The study was led by Assistant Professor Dowon Kim along with PhD Student Seongbeom Park and Professor Jinhyeok Yun from Pusan National University, South Korea. Their findings were made available online on June 12, 2026 and will be published in Volume 362, Part 4 of Ocean Engineering on July 30, 2026.
“By integrating real-time environmental conditions into navigation decisions, our system recommends optimal routes and speed profiles that reduce pollution reaching coastal communities,” explains Dr. Kim.
Unlike conventional ship-routing methods, which primarily focus on reducing fuel use or total emissions, the new framework focuses on how local weather conditions influence where the pollutants will travel. It combines sparse environmental sensor data with a physics-informed deep learning model to reconstruct high-resolution flow fields in real time. It then predicts how ship exhaust disperses under changing atmospheric conditions, and using multi-objective Bayesian optimization identifies optimal routes and speed profiles that balance environmental protection and fuel efficiency. This approach, which the researchers describe as “temporal navigation,” strategically adjusts a ship’s route and speed to take advantage of favorable meteorological windows rather than relying on uniform speed reductions.
“The greatest risk of pollution to nearby communities is often determined by when and where the pollutants are transported, rather than how much is emitted,” adds Dr. Kim.
By incorporating real-time environmental information in the ship navigation decisions, the framework enables ships to operate more intelligently and reduces their impact on public health. The researchers tested this framework in multiple navigation scenarios, including simulations based on the Busan Port vicinity. Compared with conventional navigation strategies, the AI-driven approach achieved 20–35% improvement in optimization performance while reducing peak pollutant exposure by 34%–78%.
Rather than relying on blanket speed reductions, the system adjusts a ship’s speed and route based on weather conditions to reduce pollution exposure. Its physics-informed deep operator network rapidly reconstructs airflow from limited sensor data, enabling near-real-time navigation.
Beyond improving air quality around busy ports, the framework could support digital port management, autonomous vessels, and next-generation maritime traffic systems for cleaner and more sustainable maritime operations.
Reference
Title of original paper: Physics-informed multi-objective optimization for fuel consumption and air-pollutant exposure in ship operations
Journal: Ocean Engineering
DOI: 10.1016/j.oceaneng.2026.126293
https://www.sciencedirect.com/science/article/abs/pii/S002980182602127X
About Pusan National University
Website: https://www.pusan.ac.kr/eng/Main.do
Media Contact:
Goon-Soo Kim
[email protected]
82 51 510 7928
View original content to download multimedia:https://www.prnewswire.com/news-releases/pusan-national-university-develops-an-ai-framework-for-smarter-ship-navigation-302851466.html
SOURCE Pusan National University
