StepWatch - Visualizing Pedometer Data to Motivate Physical Activity
No respectable Health and Wellness Innovation activity would address diet management without also tackling physical activity. StepWatch was a collaboration between Frank Bentley of Motorola, Cristobal Viedma of the Mobile Services Lab of Wireless@KTH, and Peter Stephenson and Adam Bell of Humana. This group dug into the Indivo X code to enhance its flexibility and reporting functionality when dealing with vital signs. They were able to store pedometer data and execute powerful queries that supported novel visualizations of step counts. Daily step goals can be difficult to manage, so the visualizations of StepWatch provide constant feedback of progress throughout the day.
This project is an exciting example of how creative researchers working on a common platform will be able to accelerate progress. During Health and Wellness Innovation 2011, this team along with a number of other MIT Media Lab sponsors and research groups came together to discuss their approaches to storing health-related data. They were able to identify commonality in their needs and began extending Indivo X in a way that would support a broader range of research. More researchers using the same platform not only means faster progress and more collaboration, but it means new opportunities for studying correlations in data.
An extension of the personal health records system, Indivo, supports rich queries of time-series data, such as pedometer or heart-rate information. By extending Indivo’s vitals document type, the system is able to support querying data from a wide range of dates, which can then be segmented in different ways, such as by hour, day of the week or month.
Furthermore, the data segments can by aggregated by summing, averaging or calculating minimums and maximums. Using this model, a watch interface (CU of the watch) can be created that allows a person to see if they are ahead of or behind their step count goals for the day.
The hour hand elongates and turns red when the user is behind their average number of steps typical for that time of day. Because the user may race to try to beat their average, the average can increase and make it harder for them to reach their daily goal. This interface is able to retrieve all the collected data in one call to the modified Indivo system to get steps segmented by hour and averaged across all days.
Another watch can show a radial progress bar of steps taken towards a goal. This means that when the user begins walking during the day, they can see the progress and determine how close they are to their goals.
These additions to Indivo are now available for others who need to store and query rich time-series data in complex ways.