Wearable Physiological Sensors: Factors Affecting Soldier Adoption & Compliance
Wearable physiological sensors can allow Army leaders to monitor the status and well-being of soldiers. The goal of this research project was to investigate factors that might impact the adoption of wearable sensor technology and soldier compliance, including: device comfort and usability; task interference; and security and privacy concerns regarding the data. The research team consisted of a USMA faculty member affiliated with Army Cyber Institute, Dr. Aryn Pyke, and two Cadet researchers, Sydney Whitham and Trey Martin (now 2LTs), who were Engineering Psychology Majors from the Department of Behavioral Science and Leadership at West Point. This research team surveyed 226 future Army officers who had used wearable sensors. Experiences included four sensor formats: on wrist (54%); on shoe (37%); and on chest strap or sports bra (9%). Shoe mounted sensors were rated highest in terms of comfort, but fail to provide key physiological data (e.g., heartrate and skin temperature). Both wrist and chest/bra sensors provide such data, but wrist sensors received higher comfort ratings (81/100 vs. 57/100). Wrist sensors were also rated relatively low for interference with daily tasks, sports, military training, and sleep (mean rating: 18/100). That said, users rated likely interference with combat at 32/100, and rated willingness to wear a wrist sensor in combat at only 61/100. One concern about sensor use was related to data access and security. Respondents rated their comfort level with leadership accessing their data to monitor their safety, divide labor, and inform performance evaluations at about 60/100. Comfort with peers having access to the data was rated much lower, at 35/100, and comfort with adversaries possibly accessing the data was rated as 9/100. If assurances could be provided that data were secure from adversaries, and that peer access could be restricted if desired, this might improve soldiers’ attitudes towards the adoption of wearable sensors.
Another factor that would facilitate sensor adoption and compliance is providing the soldiers themselves with access to the data of interest to them. The types of data that were rated of most interest by respondents were sleep metrics, overall wellness, and hydration level (mean rating: 73/100). These were followed by measures of physical and mental fatigue or stress and heartrate (mean: 67/100). Interestingly, detecting exposure to toxins was rated lower than the metrics above (57/100), possibly because it was assumed to be a rare occurrence. Skin temperature was the metric of least interest of the options listed (47/100). A few respondents also suggested other measures that might be of interest to soldiers, including: calories burned, body composition, oxygen levels and VO2 max. Thus, besides providing assurances about data privacy and security, providing soldiers - rather than just the leadership - with direct access to their data metrics of interest should facilitate adoption and compliance. How and when this information is made available to soldiers is a matter for ongoing research. Some wrist sensors do not have a display for data, which likely makes them more rugged in the field, and prevents the soldier from being unnecessarily distracted by these data. However, some of these data might be valuable to the soldier in real time, thus access to these data on demand - or in the event a threshold is triggered - would presumably be useful. This could be afforded by a display on the wrist device or another device (e.g., cell phone or heads up display). The research team would like to acknowledge and express our appreciation for the advice and information provided by Elizabeth Halford (Global Systems Engineering Contract support for CCDC), Bill Tharion (USAREIM), and Ed Argenta and Dr. Christian Whitchurch (DTRA).