Unlocking Our Personal Health Through Smart(er) Wearables
Since the introduction of the Apple Watch in 2015, the world has seen an explosion in smart watches—tiny wrist-adorning computers with the ability to track and report information on a variety of health and fitness metrics, from running speed and heart rate to blood oxygen level and sleep patterns. Consumers spend hundreds or even thousands of dollars on these wearable electronics and rely on them for myriad personal health data.
But despite how far smart wearables have come in a few short years, many limitations remain. For one, the accuracy of the health data they capture is questionable. Truly continuous monitoring also isn’t possible due to the need to regularly recharge devices. Seeing the great potential in wearable electronics, UMass Amherst researchers are working to address these shortcomings and imagine a future in which smart devices could revolutionize health and health care.
The Accuracy-Comfort Trade-off
Consumers may take for granted the data they read off their smart watches, but its accuracy is hardly unimpeachable. When it comes to electronics that monitor the body, Trisha L. Andrew, associate professor of chemistry and chemical engineering in the University of Massachusetts Amherst’s College of Natural Sciences, described an unavoidable tension between comfort and accuracy. In a clinic setting, devices such as constricting blood pressure cuffs or EKGs with five to 10 leads placed around the body, provide the most accurate readings, yet they’re not practical for people to wear around on a daily basis. On the other hand, consumer devices, like smart watches or chest straps, can report only estimates of metrics like heart rate or blood oxygen level, based on extrapolations from a single measurement point on the body.
To bridge this accuracy-comfort divide, Andrew is exploring a novel solution: smart garments, for which the entire surface area of an item of clothing could serve as the sensing element. In her Wearable Electronics Lab, Andrew has developed a technique called chemical vapor deposition to coat a thin, durable film onto fabrics. This technique can be used to give fabrics conducive, waterproof, heating, or antimicrobial properties, and even turn them into smart garments with sewn-in sensors. Using sophisticated imaging of textiles treated with this technique, Andrew’s lab has shown that the special coating wraps around each individual fiber. They have also determined that the coating is robust—able to withstand over 150 wash cycles without deteriorating—and doesn’t change the feel of the fabric.
In 2019, Andrew founded Soliyarn, a company that aims to “revolutionize the way next generation smart clothing feels and performs.” To date, Soliyarn’s primary client has been the U.S. Department of Defense, which has a particular interest in outfitting members of Special Operation Forces—who tend to operate in the most extreme climates and conditions—in smart garments. In theory, said Andrew, such garments would allow the military to monitor these individuals and react when they have reached their physiological limits. Professional sports teams have also expressed interest over the years in smart garments to monitor the physiological performance of players during games or practices.
Beyond concerns over accuracy, the usefulness of today’s wearable health monitors is constrained by the need to periodically recharge their batteries. For example, said Sunghoon Ivan Lee, assistant professor in the College of Information and Computer Sciences, many people remove their smart watches at bedtime to charge overnight, rendering the devices’ sleep tracking functions useless. Moreover, in any electronic, batteries tend to take up the most space, and they are very difficult to make flexible. The need for batteries thus stands in the way of creating smaller sensors that could be worn comfortably on the human body.
In his Advanced Human Health Analytics (AHHA) Laboratory, Lee—who has expertise in computer science, electrical engineering, and biomedical informatics—and his colleagues are working to develop solutions to these problems. Lee received the 2019 Armstrong Award to advance research on battery-less wearable sensors powered through human skin. And in June 2021, Lee, Jeremy Gummeson, assistant professor in the Department of Electrical and Computer Engineering, and Noor Mohammed, a PhD student in Lee’s lab, published a paper in the Proceedings of the ACM on Interactive Mobile, Wearable and Ubiquitous Technologies that laid out the technical groundwork for a battery-less wearable device. Called Shazam, the device uses human skin as a wire to transfer energy from an electrified daily object, such as an office desk or car steering wheel, to a sensor or network of sensors on the body. The same month their paper was published, the researchers received a grant of nearly $600,000 from the National Science Foundation (NSF) to advance development of hardware and software for this technology, including boosting its power harvesting capabilities. At present, the device is able to extract 1 milliwatt of power or less—enough to support a small, simple fitness tracker like a Fitbit without a screen, but not a more sophisticated device like a smart watch.
In the future, Lee imagines this technology would enable people to wear multiple extremely small battery-less sensors—even those that are permanently implanted like tattoos—in different spots on their bodies. These would all be charged by a single battery patch, which could take the form of a watch. Having sensors in different locations would address existing accuracy challenges and allow for collection of clinically useful data.
How far away is this imagined future? It’s difficult to predict, but given the rapid progress in wearables to date, Lee thinks it could be only 10 or 20 years from now.
“I think it may be sooner than we’d expect,” he said.
Lee envisions that the technology would first be used by people with acute or chronic health conditions, who could benefit most from precise, real-time monitoring—such as tracking blood sugar in a diabetic, body movements in stroke patients, or heart activity in a person at risk of heart failure. Further down the road, the technology could provide early warning signs to people with a family history or increased risk of certain diseases, alerting them to seek medical attention. For example, Lee thinks it could detect cancerous tumors or conditions like Parkinson’s disease (such as through a device that could alert the user to tiny tremors in their body) in very early stages, leading to quicker treatment and better results.
“This could help physicians create more personalized and patient-centered treatment programs,” Lee said. “Since the vast majority of health conditions are preventable, early detection could help people live healthier lives while relieving the economic burden on our healthcare system. Transforming health care from being reactive to proactive—therein lies the greatest potential for these devices.”
From the Lab to the Real World
As wearable technologies advance and generate enormous volumes of personal health data, they introduce a host of new questions and challenges in areas ranging from data processing to health policy to ethics.
“What do we do with all this information? Who covers the cost of monitoring? How will the information collected be tied into our healthcare system so it can be used to advance medical treatment of patients?” Lee pondered.
Andrew points out that a company like Apple, which has sold hundreds of millions of smart watches, is collecting enough data to—when combined with surveys on health conditions—eventually be useful in identifying clusters and potentially making health predictions for individuals. Andrew’s lab is currently exploring the feasibility of making such predictions, but future work would be needed to determine the accuracy of predictions and explore the ethics involved in informing consumers.
“I think the legal community will need to begin grappling with how to handle the emerging cluster of personal and medical ethics violations that are possible with this volume of data,” Andrew said.