Please add an ELEVATION/ALTITUDE field...

[AlanOfTheBerg]

Exactly! The Long/Lat precision goes way down--which is why a third dimension (altitude) would be so helpful in distinguishing among stacked roads.

Except the altitude value, which is, IMO, highly suspect in smartphones today anyway, would be off by just as much margin of error as lat/lon thereby rendering just as useable/useless.

[xteejx]

I still see some merit in this tbh. Would help a long way towards a "eco mode" for routing. Besides, GPS traces from many users can be averaged on the altitude anyway to give a decent indication of the real height in the same way it's done currently for lat/long.

[WeeeZer14]

Some info from the USGS:

* Survey grade GPS receivers typically have horizontal accuracies around 1cm. Vertical accuracy is twice that (2cm).

* Differential grade GPS receivers from 0.3 to 1.0 meter horizontal. Vertical accuracy is 2 to 3 times that.

* Consumer grade GPS receivers are between 3 and 10 meters horizontal. "This type of GPS handheld unit provides elevation data with poor accuracy."

So figure at best a consumer grade GPS is going to have 20 to 30 meter vertical accuracy. Cell phone GPS is probably also at the low end of consumer devices. And if we have stacked roads, we also are going to have signal multi-path issues. So that will make that number even larger.

Plus, GPS signals will tell us how far we are from the satellite. It will not directly tell us how high off the ground we are. There are mathematical models to estimate the surface of the earth and that is used to give an indication of where you are in relation to the ground. Different devices may use a different model or at least interpret the same model differently.

I am by no means saying this is a bad idea, I'm just saying there is a big signal to noise ratio so it isn't going to be easy.

[harling]

The Long/Lat precision goes way down--which is why a third dimension (altitude) would be so helpful in distinguishing among stacked roads.

Except the altitude value, which is, IMO, highly suspect in smartphones today anyway, would be off by just as much margin of error as lat/lon thereby rendering just as useable/useless.

Under any signal conditions, no matter how poor, even a rough estimate in the third dimension will be an improvement over NO estimate in the third dimension, when it comes to discriminating between two roads at different altitudes.

[AlanOfTheBerg]

Under any signal conditions, no matter how poor, even a rough estimate in the third dimension will be an improvement over NO estimate in the third dimension, when it comes to discriminating between two roads at different altitudes.

I disagree here. If the best case scenario is off by 20m (60ft+), that would have Waze estimating you on the bottom level of a 3-level stack instead of the top. I don't see how that is any better than providing no information at all and just using what we have. Most people do not start navigating in the middle of a stack. That and Waze deciding to re-route are about the only time when Waze should get confused about where you are in the stack. It knows how you got there, and most of these don't allow you to switch between levels in the middle, so I don't really understand why elevation is needed. The road you are on is the road Waze guided you in on, most of the time.

[xteejx]

The only reason I would see altitude data as useful is for an economical mode, less hilly terrain being more fuel friendly.
The errors we are talking about are for a single point in time by one gps device. Surely, if you have 3 or 100 at that one position, you could cut away the really high ones, the really low ones, and take the average of the rest to get a pretty accurate reading?
It could be stored in the gps trace part of the database (might even already be there), and routes calculated using that data for the economy mode as I explained above.
No one think this is a good idea, both as an idea, and technically?

Sent from my GT-I9100P using Tapatalk 2