Simulate sensor data with Health Services

Use synthetic data generated from Health Services on Wear OS to test your app as though an exercise were really happening.

If you're testing on an emulator running Wear OS 4 (API 33) or Wear OS 3 (API 30), you can use synthetic data generated by the emulator. Consult the following guide to learn more about differences between synthetic data generation for Wear OS 3 and Wear OS 4.

Use synthetic data on Wear OS 4

If you're testing on an emulator running Wear OS 4 (API 33), you can use synthetic data generated by the emulator to test your app. This introduces a number of improvements over how synthetic data is generated on Wear OS 3:

  • Synthetic data generation on Wear OS 4 is integrated with the Health Services API lifecycle. This means that there is no need for adb commands to start or stop the exercise. Instead, you can start or stop the exercise in-app as a user would.

  • Expanded support for exercise events: You can simulate receiving auto pause and resume events, fall events, sleep detection, and golf shot detection.

Use the Health Services Sensor Panel

Android Studio Koala Feature Drop (Canary) includes a Health Services on Wear OS sensor panel. The panel can be used to simulate a device having or not having specific sensor capabilities, which is important to test when building fitness apps on Wear OS.

You can also use the panel to change the values of metrics and observe how your app adjusts.

To open and use the panel, do the following:

  1. Create or open an Android Virtual Device (AVD) and run your app on the emulator.
  2. In the emulator panel, select Wear Health Services. Open Wear Health Services panel The Wear Health Services panel opens, showing a list of sensors that are available on different Android-powered devices.

After the panel opens, you can do the following:

  • Toggle among Standard capabilities, All capabilities (default), or Custom. Select Apply to send the current list of capabilities to the emulated device, and select Reset to restore the list of capabilities to their default on-off values.
  • Trigger different user events after you select the Trigger events drop-down button. From here, you can Trigger auto pause/resume of fitness activities, Trigger sleep events by the user, and Trigger golf shots that the user takes on a golf course or mini-golf course.
  • Override sensor values after you begin an exercise in an app that's installed on the emulator. After you enter new values for different exercise metrics, select Apply to sync these values with the emulator. This is useful for testing how your app handles different exercise conditions and users' fitness tendencies.

Generate synthetic data with the emulator

You can also run your app without the sensor panel and have data be automatically generated by the emulator.

Use your in-app controls to start, pause, and end synthetic data generation.

Also note that the emulator generates the same data values for each exercise.

Simulate events

You can simulate various events in the emulator such as AUTO_PAUSE_DETECTED. The following command can be used to trigger those events:

adb shell am broadcast -a "whs.event-key"

Synthetic events



Auto-Pause Detected


Auto-Resume Detected


Fall Detected


Sleep Detected


Sleep-Stop Detected


Golf Shot Detected


For example, you can use the following command to trigger an auto pause event:

​​adb shell am broadcast -a "whs.AUTO_PAUSE_DETECTED"

For golf shot events, you should specify additional parameters for golf shot swing type, which are outlined below:

Golf Shot Swing Type


Swing Putt Type


Swing Partial Type


Swing Full Type


Add the golf shot swing type after specifying the golf shot event:

adb shell am broadcast -a "whs.GOLF_SHOT" --es golf_shot_swing_type \

For example, the following command triggers a partial golf shot:

adb shell am broadcast -a "whs.GOLF_SHOT" --es golf_shot_swing_type "partial" \

Use synthetic data on Wear OS 3

If you're testing on an emulator running Wear OS 3, you can also use synthetic data to test your app.

Enable synthetic data generation

Complete the following steps to enable synthetic data generation on Wear OS 3.

  1. Enable developer options.
  2. Issue the following adb command to enable synthetic mode:

    adb shell am broadcast \

Once synthetic data generation is enabled, issue commands described on this page to control the behavior of the "synthetic user."

Disable synthetic data generation

To switch back to using real sensors, run the following command:

adb shell am broadcast \

Synthetic exercises

Health Services supports the following exercise types:

  • Walking: whs.synthetic.user.START_WALKING
  • Running: whs.synthetic.user.START_RUNNING
  • Hiking: whs.synthetic.user.START_HIKING
  • Swimming: whs.synthetic.user.START_SWIMMING
  • Running on a treadmill: whs.synthetic.user.START_RUNNING_TREADMILL

The exercises generate realistic synthetic data for the following data types:

  • Heart rate
  • Step count per minute
  • GPS location, using a single default route
  • Duration of the activity
  • Elevation and floors

In addition, the following states can be generated:

  • Sleep state—asleep or awake
  • Fall detection


To start simulating an exercise, issue the appropriate broadcast to

# start the "walking" synthetic exercise
$ adb shell am broadcast \
-a "whs.synthetic.user.START_WALKING" \

Each activity has presets for the supported metrics:

Activity Heart rate Average speed Elevation change Use location
Walking 120 bpm 1.4 m/sec 20.0 m/min true
Running 170 bpm 2.3 m/sec 20.0 m/min true
Hiking 150 bpm 1.3 m/sec 20.0 m/min true
Swimming 150 bpm 1.6 m/sec 0.0 m/min true
Running on treadmill 160 bpm 2.3 m/sec 20.0 m/min false


To stop the synthetic activity, use the following command:

adb shell am broadcast \
-a "whs.synthetic.user.STOP_EXERCISE" \


For more precise control over what metrics are generated, start a custom exercise activity using the action string whs.synthetic.user.START_EXERCISE. Provide any combination of the following flags:

  • --ei exercise_options_duration_secs <int>: duration of the exercise in seconds. Default: 0.
  • --ei exercise_options_heart_rate <int>: heart rate in beats per minute. Average: 70.
  • --ef exercise_options_average_speed <float>: average speed in meters per second. Also affects steps per minute, or cadence. Default: 0.
  • --ez exercise_options_use_location <boolean>: whether to emit location data during the exercise, using a default route. Default: false.
  • --ef exercise_options_max_elevation_rate <float>: maximum possible elevation change rate in meters per minute. Default: 0.

For example, set exercise options in the following way:

adb shell am broadcast \
-a "whs.synthetic.user.START_EXERCISE" \
--ei exercise_options_heart_rate 90 \
--ef exercise_options_average_speed 1.2 \
--ez exercise_options_use_location true \

You can also change the data types that are available, whether or not the real or emulated hardware supports a particular data type. For example, you can enable or disable absolute elevation, as shown in the following snippet:

# enable synthetic mode and enable absolute elevation
$ adb shell am broadcast \
--ez absolute_elevation true \

# enable synthetic mode and disable absolute elevation
$ adb shell am broadcast \
--ez absolute_elevation false \

Other states and events

Sleep state

You can also trigger sleep states for the synthetic user. Two states are supported: asleep and awake.

To enter the asleep state, run this command:

adb shell am broadcast \
-a "whs.synthetic.user.START_SLEEPING" \

To enter the awake state, run this command:

adb shell am broadcast \
-a "whs.synthetic.user.STOP_SLEEPING" \

Fall detection

To simulate a fall, run this command:

adb shell am broadcast \
-a "whs.synthetic.user.FALL_OVER" \

It can take up to a minute for health services to deliver the fall event.