The aim of this video is to provide our experience from deploying wearable sensors embedded in commercial smartwatches for the proposals of the live media project, a population study focusing on the mitigation of health effect. The project involved two distinct population groups patient with atrial fibrillation, asthmatic children aged six to 11 years old for the assessment of their physical activity and their personal location. Both groups were equipped with a commercial smartwatch.
The participants were required to wear the smartwatch daily and data were transmitted over their home's wireless network to a collection platform for almost realtime assessment of compliance and data quality. In this video, we displayed a simple, cost-effective and off-the-shelf solutions that we implemented in the field to improve user compliance, data quality, and data completeness. In addition, we showcased data completeness improvements from of such protocols using the representative results from the live media study.
Launch the device automation application and navigate to the interface complete tasks. Add a macro or a task by clicking the plus sign. Click the plus sign on the trigger tab.
Select Date and Time. Select the Regular Interval tab. Don't use the reference time.
Set the regular interval. Set the fixed time interval to act as a trigger to initiate the procedure. Click OK.To add an action, click the plus sign on the action tab.
Select the Conditions/Loops tab. Select close. Add a condition by clicking the plus sign.
Select the Date/Time tab. Select Stopwatch. Click OK.Set a fixed duration to be added as a condition.
As part of the procedure, the condition being tested is whether a stopwatch is either more than 29 minutes, or less than one second. The former reflects our preference to systematically activate the data collection application at least every 30 minutes. And the latter represents a case that the stopwatch is not working or has stopped.
Change the end by clicking on the triangle. Select Or.Whenever one of the condition is met, the process can continue. To add an action between the statement, click on the End If.Select Add Action above.
Navigate and find the Screen tab. Select the Action Screen On and click OK.When the conditions are met the tasker will enable the screen. Following similar steps at the second action that reduces the brightness of the screen to the lowest percentage.
Add an additional action that resets and restarts the stopwatch that is being used in this automated process. This creates a loop. Follow similar steps and add an action that enables collection application.
Select Applications. Select Launch Applications. Find and select the Data Collection application.
Select Force new and click OK.Our process is complete. The trigger, the conditions, and the actions are set. Give a name to this process and select SAVE.
Create a process that systematically enables the wifi of the smartwatch. Add a trigger to create a start point for the process. Select Device Events and select Screen On off.
Select Screen On and click OK.Add an action by clicking the plus sign on the Action tab. Select the Conditions/Loops tab. Select the close.
Add a condition by clicking the plus sign. Select the Connectivity tab. select Wifi State and choose Wifi Disabled.
Click OK.Select OK.Click on the End If.Select Add Actionable. Select the Connectivity tab. Click on the Wifi Configured tab.
Select Enable Wifi and click OK.Upon screen activation, and if the condition is met, the process activates the wifi. Add an additional action between the If statement that resets and restarts Stopwatch Two. This stopwatch is important for disabling the wifi after a few minutes.
An action that is carried out by a different process in the protocol. Give a name to this process and select SAVE. Create a process that systematically locks event information.
Click the plus sign on the Trigger tab. Select Date and Time. Select the regular interval.
Don't use the reference time. Set the interval at five minutes and click OK.To add an action, click the plus sign On the action tab. Select Logging.
Click on Log Event. Click the Ellipses Points tab. Find and select Wifi SSID.
Select OK.By clicking again on the Ellipses Point tab and following the exact same steps, you can further add Wifi signal strength, device serial number, GPS coordinates, accuracy of GPS signal, as well as whether the smartwatch is charging or not. This action creates a log event with pre-specified variables relevant to the project. Give a name to this process and select SAVE.
Before the smartwatch is given to a participant several important settings must be modified. Swipe the smartwatch screen to the left and navigate to the smartwatch settings function. Click on Settings.
Select Sound, and disable all sounds by minimizing the volume levels of all individual settings. Swipe the screen to the right to return to the main setting screen. Scroll down and select the Connect function.
Select GPS. Click on mode and set the GPS setting to high accuracy. Swipe the screen to the right to return to the main setting screen.
Scroll down and select power saving function. Deactivate Standby intelligent power setting, and check that the battery saver is always turned off. Swipe the screen to the right to return to the main setting screen.
Scroll down and select Date Time function. Disable the Automatic time zone tab. Scroll down and click on Select time zone and select the correct time zone and enable the 24-hour format.
Scroll up and enable the Automatic time zone and check that the Automatic date time mode is set to use network-provided time. Swipe the screen to the right to return to the main setting screen. Scroll down and select More.
Open the option Background Cleaner and disable the Battery Saver. Swipe the screen to the right to return to the More settings screen. Scroll down and select App Freeze.
Click on Next. Scroll down and click on Google Play Store. Select Freeze.
The app will automatically disable. Swipe the screen to the right to return to the More setting screen. Select Notifications.
Select Google and block all notifications from this app. Depending on the device and operating system, other applications may also pop up notifications. Swipe the screen to the right to return to the More setting screen.
Scroll down and select Data save. Disable the Data save. Swipe the screen to the right to return to the More setting screen.
Select App settings. Scroll down and select Special access. Click on Battery optimization setting.
Click on the triangle and select All apps. Scroll down and find the application locker and select Don't optimize. Find all relevant apps that are used in the smartwatch setup and select Don't optimize.
The main applications that perform or support a must operate without any battery restrictions. After you finish with the setup, perform a check up of the settings that we have done in this step. Check again that you have minimized all sounds and vibrations, that the GPS mode is in high accuracy.
Date and time settings are correct, and furthermore, make sure to disable any settings that interfere with the smooth operation of the apps in the background. Additionally, make sure that you have freezed any apps that may perform updates. Make sure that the main applications that perform or support a are operating without any battery restrictions.
Participants were given specific instructions to charge the smartwatch every night and wear it during the day. The configuration of the device displayed in the protocol in combination with the automated procedures developed with the tasker, minimize any extra hassle for the participant. Here we provide exemplary results from a total of 17 participants from the asthma panel during spring 2020.
The participants were equipped with the smartwatch and providing data two weeks before and after protocol implementation. Figure 7A presents a percentage of time with data for each participant separately, while Figure 7B presents the corresponding distributions of percentage of time with data for the whole group. An increase in percentage of time with data was observed.
Protocol implementation led to a statistically significant increase in data completeness of approximately 15%Furthermore, Figure 8 and Figure 9 presents two different dates from a single AF patient participating in the live media study in spring 2020. Figure 8A and 9A presents the actual raw GPS signals. In Figure 8B and 9B, we have the implementation of the GPS data filling algorithm that allows replacing missing data with estimated values.
Figure 8C and 9C present signals from the smartwatch connectivity with the Wi-Fi network. The algorithm performed reasonably well in Figure 8A compared with the network signals in 8C. However, in Figure 9, the algorithm alone did not accurately estimate all missing data, as it failed to capture that the participant also returned home for a period of about 90 minutes.
The inclusion of a Wi-Fi connecting signal, as collected by the tasker application, can significantly prove the performance of the algorithm and minimize misclassification. This protocol was employed in spring 2020 and was used to set up smartwatches that were given to asthmatic children enrolled in the live media project. As a result, the children were equipped with the smartwatches during the first wave of the COVID 19 pandemic and this gave us the opportunity to objectively assess their compliance to COVID 19 lockdown measures.
The results of that work were published previously Scientific Reports General and demonstrated that with escalating lockdown measures children increase their daily fraction of time spent at home and reduced their physical activity in both countries. In summary, this work provides a protocol which includes simple and cost-effective solutions to real life challenges in population studies employing consumer wearable devices. This approach can be easily replicated or adapted by researchers working with wearable devices in the fields of clinical research and public health.
