Sleep Cycle App
TL;DR β¨
More than six months ago, the author decided to try the Sleep Cycle application. It primarily functions as an alarm clock: the application is started and the phone is placed under the pillow. It communicates with the gyroscope and maps sleep stages based on the amount of body movement during the night. It wakes the user when the right moment arrives.
After several months of collecting data, the author presents clear statistics for review. The information was processed and plotted by the program itself.
1. Introduction¶
More than six months ago, the author decided to try the Sleep Cycle application. It primarily functions as an alarm clock: the application is started and the phone is placed under the pillow. It communicates with the gyroscope and maps sleep stages based on the amount of body movement during the night. It wakes the user when the right moment arrives. As a smart application, it recognizes that few people wake at an exactly defined time; everyone has a biological interval during which they are almost awake. The application uses this interval and waits for the most suitable moment (the REM stage) to wake the user.
After several months of collecting data, the author presents clear statistics for review. The information was processed and plotted by the program itself.
2. Data¶
The application is relatively simple and processes only the two most important pieces of information: sleep quality and sleep duration. It compares these two parameters with other user inputs. The following images visualize the data:
Figure 1: Every morning, the software plots a summary for the day as a graph showing the alternating sleep stages.
Figure 2: The most interesting trends¶
Figure 3: A visualization of the time at which the application was started before sleep.
Figure 4: The trend in average sleep duration over the months.
Figure 5: The trend in average sleep quality over time.
Figure 6: Correlation between activities and sleep quality over the entire observation period.
Figure 7: Average sleep quality over the course of a week, calculated as the average for each day across the entire observation period.
Figure 8: Average sleep duration, calculated as the average for each day across the entire observation period.
3. Commentary¶
Sleep quality is a rationally defined parameter, but it does not reflect reality. It is often distorted when sleep duration is insufficient and there are many REM stages. This produces a false positive that reduces the information value in the short term.
Figure 6 shows a nonsensical correlation between sleep quality and the recorded activity. The author physically felt rested and well slept after working out at the gym or running, but the application assessed the sleep quality as insufficient because there were too few REM stages.
Because its measurements rely on a gyroscope, the application cannot assess various physiological disturbances, such as deviations in sleep caused by alcohol consumption. Here again, an excessive number of REM stages and a short sleep duration produce false-positive results.
The application also provides many additional features that are not described here:
- Setting a pleasant melody for waking up in the morning.
- Measuring heart rate after waking up (the measurement uses infrared radiation from the thumb).
- Adding a smiley that records how the user feels after waking up.
- Playing a pleasant sound before bed to make it easier to fall asleep.
- Exporting raw data in CSV format for independent processing.
4. Conclusion¶
Overall, the application is good for entertainment, but its technical shortcomings prevent it from reaching a mass audience. It displays distorted information, though some of these distortions can be eliminated through long-term observation and use of the application.
The application faces a genuine risk of being pushed out of the market by competitors such as Fitbit or Jawbone UP, which can also measure heart rate and are more convenient to use.