Stand Hours Algorithm Needs Changing

This is a user-to-user forum. No one here (with VERY limited exceptions) works for Apple. If you would like to make this comment to Apple, use this link:

Feedback - Watch - Apple

I have been writing software for decades, but I started with a BSEE degree, because it was before a Computer Science degree was common in most universities.

And yes, having an engineering degree, as well as being a software developer, means I sometimes can be anal about details.

But, trying to capture the real world (Standing each hour), while staying inside a very constrained power budget, with minimal sensors, makes for a very complex algorithm, that is going to have holes in it you could drive a semi-truck through.

Sitting at my desk typing generates tiny accelerations. How do you distinguish those from walking around, or standing? You cannot be running the GPS radio all the time, because it blows your power budget, and indoors it is not sufficiently accurate. The altimeter is not accurate enough to register a move from sitting to standing, so the Apple Watch cannot really use that. And even if you invent a better battery, you are still limited on power, because most of the power used results in waste heat, so spending power faster, creates more heat, and the Apple Watch is sitting on skin, that does not like 3rd degree burns.

As it is, choosing to look for arm motion, that might indicate walking, is a conceptually simple algorithm, and if detected, has a reasonable chance of being right (unless you knit, then it is totally wrong, but you get huge daily step counts 😁).

All I can suggest is that since you are motivated, and you are an engineer, create and test the algorithm that works in most real world situations for lots of people, does not drain the battery, does not cause skin burns from excess heat, then give it to Apple. Motivated and inventive people are why there as so many apps for smartphones. This is an opportunity to fill a gap. Be that person.

Another option is to do a one minute Workout. The Workout is Other > Open. Let it run for just 1 minute and stop the workout. Making the perfect algorithm is impossible, no two people are alike physiologically, so one algorithm won’t work for everyone.

At the end of the day, it’s just a stand goal. If you know you completed the goal, isn’t that enough? 😀😀😀

How do you suggest Apple Watch measure "Moving Around"? If you are not going to use the accelerometer, then that leaves GPS.

https://mapscaping.com/how-accurate-is-gps/#:~:text=With%20a%20clear%20view%20of,(33%20to%2066%20feet).

With a clear view of the sky, most consumer-grade GPS receivers can determine your position with an accuracy of about 3 to 5 meters (10 to 16 feet). In real-world conditions, such as when there are obstacles like buildings or trees, the accuracy can decrease to about 10 to 20 meters (33 to 66 feet).Sep 16, 2023

So if you are in your home, or say you are a chef in a kitchen, all your moving around might report the exact same 10-20 meters, so no movement at all.

Sure it would work kind-of-OK on long walks in a mostly straight line, although then you need to know the stride of the user so you know if you should count 1, 2 or 3 steps per meter.

I am not saying it cannot be solved, but I think your assumptions are a bit simplistic, and the algorithm is going to be rather complex.

Also running the GPS 24/7 will chew up a lot of battery, which is why it is used infrequently, unless you are running one of the exercise modes (walking, running, bicycling, etc...).

Also does an accelerometer measure lateral motion, or does it measure accelerometer and deceleration?

Going a stead speed, whether it is 1 meter/second, or 26 meter/second should not apply any force to an accelerometer,

After all the Earth is rotating at 1600 KM/hour, and the Earth is orbiting around the sun at 107,000 KM/hour, and the sun is orbiting around the galaxy at 828,000 KM/hour. All of that is lateral motion, and I do not think the accelerator in the Apple Watch is compensating for all that steady state motion depending on which direction you are facing.

Then again, I have never placed with a solid state accelerometer, so I'm not sure what the outputs are.

So I'm driving/riding in a car and thus I have lateral motion.

And how many steps to count per minute, since you do not know how fast you are going, or is that recorded by lateral motion?