Fitbit+: A behavior-based intervention system to reduce sedentary behavior

Abstract

Self-tracking wearable devices are capable of tracking calorie consumption and inferring physical activity physical activity to support self-awareness and healthy behavior. These devices automatically capture human behavior (such as walking) but do not typically make the user aware detected unhealthy behaviors. Furthermore, these devices cannot intervene in the moment to make users aware they are engaging in unhealthy behavior (such as sitting for a long period of time) and persuade them to correct these unhealthy behaviors (e.g., by taking a break to go for a walk). There is an increasing trend for people with low physical activity occupations to sit for long periods of time, yet research suggests that lengthy sitting, independent of overall physical activity level, increases the risk of weight gain and mortality [1]. We aim to decrease the duration of sedentary bouts in the workplace by detecting when people have been inactive for a long time and then prompting them take a short break from their desks. In this poster we present the design of Fitbit+, a system that realizes this strategy by leveraging Fitbit's near real-time, automated step logging to detect sedentary behavior and then prompt users to take short breaks.