Apple Watch on Omnidirectional Treadmill

My KatVR omnidirectional treadmill does not calculate fitness metrics. And my pedometer fitness sensor (which goes in my shoe) would go to sleep if I stopped walking for more than a few seconds, and also would take like 30 seconds of continuous walking to “wake up”. Playing a videogame is not continuous walking, it is a lot of “start and stop”. So I would do 3 hours or so on the omnidirectional treadmill, sometimes walking or running and sometimes standing still, and the pedometer would only record like 20 calories burned and 0.01 miles traveled, or some absurdly incorrect number like that. Therefore, the only device I have which would successfully calculate fitness metrics while using my omnidirectional treadmill was my Apple Watch.

I found that playing Fallout 4 VR on my omnidirectional treadmill would, according to the Apple Watch, result in 0.3 to 1.5 miles walked in 3 or 4 hours of gameplay. I would have to intentionally walk long distances in the game (instead of “fast traveling”) if I wanted to walk more than a mile (according to my Apple Watch). Also, when indoors in the game and during battles, I would usually stand still and move around in the game using my handheld controllers because this was more convenient in the game. This is because 1) there is a slight lag in the omnidirectional treadmill motion being entered into the game, 2) rotating my body on the omnidirectional treadmill was slow because your body rotation has to physically push the rotating back support structure of the omnidirectional treadmill, and 3) walking slow is very difficult on the omnidirectional treadmill due to having to overcome the slipping resistance of the special shoes against the surface of the omnidirectional treadmill.

Note that I modded the Fallout 4 VR game to make walking/running movement 4 times faster in the game. This is because I found when I was running on the omnidirectional treadmill, my character would still be moving tediously slowly in the game. There is a speed multiplier setting in the omnidirectional treadmill configuration application, but even with that setting at max speed, it was still too slow (in my opinion) in the game. If I hadn’t modded the game to make my character move faster, slow waking in the game might’ve been more practical.

I have not tried to play any games other than Fallout 4 VR on the omnidirectional treadmill. The amount of walking or running would presumably vary a lot depending on the game you play.

My guess is that the distance traveled according to the Apple Watch on the omnidirectional treadmill is fairly accurate. The calorie count (according to the Nike Run Club app on the Apple Watch) seems to be way too high, just like it was on the normal treadmill. For example, if walking 1 mile in 4 hours on the omnidirectional treadmill, it would estimate that I burned like 500 calories. I suppose walking a mile on the omnidirectional treadmill is more exercise than 1 mile in real life, because you’re using extra energy to overcome friction when sliding the shoes over the treadmill surface, which kind of feels a little like you’re walking through mud or shallow water. And you’re also using your arms and such. But 500 calories in this example still just seems too high.

Apple Watch vs. Other Fitness Sensors on Exercise Bike

TBD

Apple Watch vs. Other Fitness Sensors on Rowing Machine

TBD

Apple Watch vs. Other Fitness Sensors on Treadmill

I did an experiment by exercising on a treadmill and comparing the results calculated simultaneously by the following items:

1) heart rate, distance, and calories counted by the Apple Watch and displayed by the Nike Run Club watch app (which is integrated with Apple Health)

2) heart rate from a Wahoo Tickr heart rate sensor (worn around your chest, typically used by cyclists) the distance from a Polar Run pedometer (worn on your shoe), and calories calculated by the Zwift PC app, as displayed by the Zwift PC app

3) the distance and calories calculated by my treadmill

Note that the Apple Watch and Zwift (1 and 2 above) had my age, weight, and height programmed into them, but I believe my treadmill (3 above) did not

My results, as shown in the video below, were that 1) and 2) calculated the heart rate and distance within about 95% accuracy of each other. The distance calculated by 1) and 2) was maybe 15% lower than 3), but I believe that may be partially from treadmill tread slippage or something like that, so I consider the distance calculated by 1) and 2) to be fairly accurate. At some point in the future, I should go out into the real world with these sensors, run a known distance, and double-check the accuracy of these distance calculations.

The calories calculated by 1) 2) and 3) were widely different. 1) calculated twice as many calories as 2), and 3) was somewhere in between. I believe that the actual calories burned during the workout is probably most accurately determined by 2), which is unfortunately the lowest number, primarily because the pedometer in my shoe in setup 2) can more accurately measure my running stride (or more accurately, my speed-running stride) on the treadmill. Also, the Zwift app is for serious cyclists and runners who care deeply about fitness metrics, whereas the Apple Watch is more for casual fitness, and I’ve seen through experience over the years that casual fitness devices almost always exaggerate calories burned, sometimes to a comedically ridiculous extent

Note: The heart rate sensor of setup 2) failed to connect for about 3 minutes, and this may have caused the calorie count of setup 2) to be slightly lower (but I still think the Apple Watch is massively over-counting the calories, though)