Say you're a ship captain and you want to know when one of your sailors is fatigued to the point he or she is at risk of getting injured.

Some University of Connecticut students have an idea.

The team of Devon Thompson, Julia Podsen, Yannis Halkiadakis, Prateek Rana and Kyle-Gabriel de Vera Tan, has developed a device, worn around the ankle, that detects normal and abnormal walking patterns. The goal is to develop an algorithm that can predict from someone’s walking patterns if they’re going to get injured and what the injury will be.

 “The whole idea is, 'Can we keep these people mission-ready? Can we keep them performing optimally on a boat?' ” said Kristin Morgan, an assistant professor in UConn’s Biomedical Engineering Department.

The sponsor for the project is Electric Boat, which along with the Navy has made a big push to use wearable technology, given its low cost and ease of use, to monitor human performance.

A year and a half ago, UConn and the University of Rhode Island teamed up to create the Naval Science and Technology Program to get students interested in naval projects before they graduate, and to help create a pipeline of workers for the naval industries in the two states. More than 100 students have been involved in the program since it started.

Graduating engineering majors at both universities have to complete a senior project — that amounts to about 500 projects total between both universities. This year, 17 of the projects have naval relevance. The goal is to increase that number next year, and include some of the smaller suppliers of the naval industry in both states.

The wearable device project at UConn is just one example of what the students have been working on. As part of the project, the students have observed different subjects — some who are healthy, others who have lower-extremity injuries — going through the same walking protocol on a split-belt treadmill, which allows them to move their feet at different speeds. That destabilizes the subjects to see if they can restabilize. They also observed them at symmetrical speeds.

The subjects wear the sensors around each of their ankles. The sensors have a small microphone attached, which capture the vibrations of their footsteps, which are then used to differentiate between abnormal and normal walking patterns.

The sensors also have a Bluetooth device attached, allowing them to collect the data in real time, which is plotted on a graph on a computer. The next step is to put the data into an algorithm that can be used as an early diagnostic tool to know if someone is susceptible to injury.

The students said there’s a myriad of uses for the tool such as sports rehabilitation. If you can predict when someone is going to get injured, and the type of injury, you could give them exercises for example to help prevent that from happening.

j.bergman@theday.com