Locomotion tool testers (Cardboard/GVR/Daydream) needed

Dear fellow VR game developers,

I’ve been working on locomotion tool for Cardboard/GVR/Daydream for several weeks already and it works great for me, but I need some external tests to make sure it works for others as well.

Here is some demo video:

(the blue cord is an audio cable since Android doesn’t allow to record audio from device directly)

Current features include:

  • Step detection (with leg distinction)
  • Step rate computation
  • Jump detection
  • Crouch detection
  • Evade move detection (left/right)

I’ve prepared simplistic demo app (Android 5.1+) for test (which includes first 3 features) and ask you to download it and write here some feedback. The link for app is

The demo setup can be seen on this picture:
Demo setup picture

App includes 2 different step detectors (based on accelerometer X and Y data), which can be switched by tapping the screen.

The instructions are following:

  • Download and install app
  • Run app and try walking in place. You should hear the voice, indicating which leg have you used for the previous step. If you don’t hear it - try increasing volume or switch step sensor type (by tapping screen)
  • Try jumping and moving around the level
  • If detection seems off too much, try switching the sensor type
  • Please write your feedback (what was the overall feeling? how accurate was leg detection and distinction?) here and (if possible) attach log file(s) located in /android/data/com.Kelt.BMDetectorTest/files/BMDebugLogs. Logs contain only accelerometer data and derived variables (you can check that yourself)

If all will be ok, I’ll publish the tool (BodyMovementDetector) later in the Asset Store for really affordable price. (the most active testers will get it for free from me after that)

Let’s try to improve VR experience together. Locomotion can really bring even more than room scale even for mobile, which (for me, at least) feels really inspiring.

1 Like

I like it, making the best of what’s available signal wise.

You can also look for fast impulse in accelX to detect user tapping HMD side, I tried that and it works fairly reliably.

And can analyse microphone signal with FFT for fingersnap or handclap detection to fire a weapon at gaze point.

I have not had time to try your demo yet, but I will…