A research team working in North Carolina State University recently came up with an energy-efficient technique that helps in accurate tracking of a physical activity on the basis of data retrieved from wearables devices. One of the major goals of this health technology is to trade-off between power and accuracy required for storage and data assessment. The last two pose big challenge as wearable devices have very limited power.
According to Edgar Lobaton, an electrical and computer engineering assistant professor at NC State and also the senior author of this paper, “Tracking physical activity is important because it is a key component for placing other health data in context. For example, a spike in heart rate is normal when exercising, but can be an indicator of health problems in other circumstances.” Any technology developed for keeping a check on physical activities needs to face two major challenges. The first one is that the program needs to estimate the data that is to be processed whenever a physical activity takes place. For instance, the data collected for a 10-second increment needs double computing power as compared to evaluation of data over five-second activity or tau.
The next challenge is to store this information. A simple answer to this problem is to pile up similar actions profile in one tie up under single heading. Like, you may group together specific data signatures under one heading called “running” while a few others may be lumped or piled up as “striding” or “walking”. The main problem is to find a single formula that permits the program to recognize meaningful profiles. If the formula turns out to be a very casual one, the profiles will become too broad and completely meaningless. But if the formula is too categoric, you may have multiple activity profiles that are tough to be stored as relevant information. To find solution to this problem, the researchers came together and did five activities together: sitting, waving, walking, biking, and golfing. They then assessed the resulting data with taus of zero seconds, then two seconds then four seconds and so forth till 40 seconds. Then they experimented with various parameters for classification of activity data into particular profiles.
Lobaton further adds, “Based on this specific set of experimental data, we found that we could accurately identify the five relevant activities using a tau of six seconds. This means we could identify activities and store related data efficiently. This is a proof-of-concept study, and we’re in the process of determining how well this approach would work using more real-world data. However, we’re optimistic that this approach will give us the best opportunity to track and record physical activity data in a practical way that provides meaningful information to users of wearable health monitoring devices.”
Motion capture technology helped researchers at NC State University develop an energy-efficient technique for accurately tracking a user’s physical activity based on data from wearable devices. (Image courtesy of Edgar Lobaton.)
Filed Under: News