The vehicle correlates elevation raises and drops measured against their location via due north to create yet another metric to navigate with.

So the unit takes a sample, it's at 25 feet above sea level. After driving 2 minutes at 60 miles per hour north plus 23 degrees it's now at 48 feet above sea level.

By the time you have a 4th or 5th sample you've drawn a pretty distinct terrain pattern and can make a pretty good assumption of where you are.

This can be done with barometric pressure readings or accelerometers measuring up and down motion the same as exercise wrist bands telling you if you've gone up or down stairs.
-- doctorremulac3, Jan 07 2017

You're going to need some pretty sensitive altimetry to pinpoint your height to within, say, 10 feet. If you use a pressure-based altimeter, the variation due to barometric variations will be way bigger than that.

Of course, you could always get your height from the GPS...
-- MaxwellBuchanan, Jan 07 2017


Yes, but the relative elevations between 3 or 4 points wouldn't change due to weather conditions and that's what you'd be measuring.

Even apparently flat landscape is pretty lumpy and it's sort of like Earth fingerprints. You get a big enough group of sample points with a particular compass location it's a pretty safe bet you'd know where you were.

I think electronic barometers are pretty sensitive.

But fine, measure the variation between points with accelerometers which is how I think they have those exercise measurement wristbands tell if you're going up the stairs.

So here's how it words. You're driving from point A to point B and you turn this thing on. It says "Ok, you're at level X." Let's start measuring your up and down / speed / compass location and draw a 3d line. Doesn't matter what the starting elevation is. It says: "Compass location Y, 60 miles per hour to Y1 after 5 minutes elevation is X+2, after 5 more minutes Y2, X-4," etc till you have a nice little unique picture of this particular path, you compare that 3d path to the 3d paths in the database and find your location.

I suppose you could even measure tilt of the vehicle. Tilted down 2 degrees and drove for 5 minutes then tilted up 5 degrees for 3 minutes etc.

So yea, barometer might be a bit sketchy but there are plenty of other ways you could do this.
-- doctorremulac3, Jan 07 2017


You're probably right.
-- MaxwellBuchanan, Jan 07 2017


Relative altitude? Just measuring the car going up and down? Ah, I get it. Big potholes.

It's all about the potholes, I tell you!
-- mitxela, Jan 07 2017


Hang on. You'd need a correction factor as global warming -> sea level rise, and everything's height is measured against sea level.

Also there's an upper limit, as according to my calculations on the 22nd of January 2089 - at about tea-time - the very tip of Mt Everest will be finally and permanently under the waves.

And then where will be, eh eh? Tell me that!
-- not_morrison_rm, Jan 08 2017


I'm steering more towards the "angle of car / attitude determined by accelerometer / speed calculation / relationship to due north /" model as being impervious to such issues as apocalyptic weather changes and such.

That being said, by the time Mt Everest is under water all my designs are no longer under warranty, expressed or implied.
-- doctorremulac3, Jan 08 2017


//You'd need a correction factor as global warming -> sea level rise//

I believe the factor is 3.3mm per year, and has been since 1880 (and probably before that). The sea is smarter than we give it credit for: it started rising in anticipation of anthropogenic CO2 emissions.
-- MaxwellBuchanan, Jan 08 2017



random, halfbakery