AI system could help protect health of US Navy divers

The research team at the University of Illinois Chicago (UIC) was awarded a grant of $725,000 (£520,000) by the US Office of Naval Research to pursue the project.

Sailors are sent into all kinds of water as part of their service in the US Navy, but they have limited resources to understand in real-time the health risks that may exist when they conduct underwater duties – from fleet maintenance and repairs to search and rescue and research missions.

The most reliable water testing technologies typically rely on lab-based analysis of samples and scientists knowing which microbes to screen. But with dynamic weather, currents, water temperatures, and sewage and pollution factors, the exact condition of water, particularly of coastal water, at a specific time is hard to predict.

“By the time a water sample arrives at a lab and is tested, the conditions may have changed,” said Dr Samuel Dorevitch, associate professor of environmental and occupational health sciences at the School of Public Health. “If Navy divers had real-time information, they could select the best protective equipment, dive duration and take other measures to prevent the various health issues, like heat stress or gastrointestinal, skin, and respiratory infections that may result from microbes in water.”

The researchers on the project believe that a novel approach using artificial intelligence can make a difference in tackling this widespread issue.

“Artificial intelligence offers a way to synthesise a vast amount of information quickly for a specific calculation, and this technology, if we can bring it to fruition, provides an opportunity for us to improve the tools available to the Navy,” said Isabel Cruz, a professor of computer science at the College of Engineering.

The researchers hope they can develop a system that can be used in any location by divers to analyse water conditions through a combination of user-provided and web-based information and human data, such as the age of the divers, their health, and the size of the diving team.

“This project is both exciting and challenging because of its multidimensionality,” said Cruz. “We hope to pull information from many sources that offer different data, and we will have to integrate data that are quite complex, heterogeneous, and often without metadata.”

Cruz said that the team intends to build the AI and machine learning methods in stages. “If we can teach our system to reliably and accurately filter and prioritise all these data for risk prediction, I think we will have something remarkable,” she continued.

Meanwhile, Dorevitch added that if the team could provide divers or their commanders with a handheld device or app to evaluate the ever-changing ecosystem of a particular body of water and any potential health risks at the time they enter the water, parties involved would be better able to plan their mission for optimal health and safety.

“For those in the Navy, getting in the water is not optional and anything we can do to aid quick, data-driven decision-making for mitigating health risk is beneficial,” he concluded.