​Devices for Activity and Health Tracking

Today, there are an increasing number of activity and health tracking devices in our lives. How do we leverage their overlapping capabilities, and inform the design of the next generation of such devices?

SC Faculty and Researchers

​Mayank Goel

In recent years, health sensing on mobile devices has received considerable attention. This can primarily be attributed to their ubiquity, lowering costs, and computation and sensing capabilities. A key trend emerging from the popularity of mobile devices is the quantified self movement. Devices for fitness, such as FitBit, heart rate monitoring on smartwatches, and other vital signs, are becoming increasingly common. These systems demonstrate how a mobile device, with all its computational capabilities, is an excellent candidate for health and fitness tracking. As more and more of consumer computing devices gain health-tracking (or general sensing) capabilities, three questions emerge:

  1. How do we use these devices for health sensing?
    Barring a few exceptions most health tracking devices have an external sensor at its core. For example, there are a number of Bluetooth-enabled blood pressure monitors in the market that stream the measurements to a connected smartphone. Although these systems are very useful to keep track of various physiological parameters, they do not improve access to healthcare. The user still needs to carry an extra device for tracking their health. While for a user that can afford this extra device there is an inconvenience of buying, maintaining, and carrying this device, the implications are serious for the user who cannot afford this extra device. For example, adding another piece of hardware makes the technology less accessible to the people in the poorest regions of the world. 78% of the world’s 7 billion mobile phone subscriptions are in developing countries and almost 10% of those phones are smartphones. Therefore, if we only use the sensors that are already present on mobile devices, we immediately have an opportunity to convert billions of phones into medical devices.

  2. How do we modify these devices for better health sensing?
    As we leverage the capabilities of current mobile devices, we gain a valuable insight into their limitations. While the current sensors might be useful for many impactful applications, there will be many more applications where the sensors will prove inadequate. An important research questions then becomes, how we do modify these devices for better health sensing? In many cases the modifications might be as simple as adding a new LED to the phone. Studying these modifications in detail and helping the sensors evolve can be beneficial.

  3. What new devices do we make?
    As we push the current devices to their limits, we also have an unprecedented opportunity to inform the design of next generation health devices. If one compares the hardware capabilities of a smartphone, a tablet, and a smartwatch, there are significant overlaps. The biggest differentiating factor is the form factor, and that can have a big impact on a device's capabilities. Evaluating the design of the next generation devices and investigating various form factors can help in understanding the advantages of various approaches. For example, a FitBit on the user's wrist is quite accurate for step counting and heart rate estimation. However if the user has multiple such devices on their body (may be one on their chest and another one on their feet), the estimates might be even more accurate.


Project Publications

Mayank Goel, Elliot Saba, Maia Stiber, Eric Whitmire, Josh Fromm, Eric Larson, Gaetano Borriello, Shwetak Patel
SpiroCall: Measuring Lung Function over a Phone Call

Mayank Goel, Eric Larson, Gaetano Borriello, Sonya Heltshe, Magaret Rosenfeld, Shwetak Patel
SpiroSmart: Using a Microphone to Measure Lung Function on a Mobile Phone

Mayank Goel, Eric Whitmire, Alex Mariakakis, Scott Saponas, Neel Joshi, Dan Morris, Brian Guenter, Marcel Gavriliu, Gaetano Borriello, Shwetak Patel
HyperCam: Hyperspectral Imaging for Ubiquitous Computing Applications

Lilian deGreef, Mayank Goel, Min Joon Seo, Eric Larson, James Stout, James Taylor, Shwetak Patel
BiliCam: Using Mobile Phones to Monitor Newborn Jaundice

  • Related Research ​IoT Expedition Read More
  • Related Research ​Personalized Privacy Assistant Read More