I've been playing with an Android-based app on my phone to collect some data about my riding. The goal is to see how consistent I am from lap to lap. In addition to the sensors (that include accelerometers and gyroscopes) and the internal GPS it connects to an external GPS via bluetooth. I did this because, quite honestly, most phone GPSes stink for this kind of thing. They update once per second (at most) and are accurate to maybe 20ft. I picked up an external GPS for about $90 that updates much faster and with greater accuracy.
The app isn't nearly done but I took it on my drive yesterday to collect a large data set. Here's the position and elevation data it gathered.
Which correlates well with the Google-Map route I posted yesterday. So my code isn't completely broken. (The two gaps in the elevation graph are the two long stops I made for food 9:15 and 12:30.)
But since I'm not planning to use this 460-mile loop as my track I need to look at the noise at a much smaller scale. I found a nice hairpin turn to focus on. (The little red circle in the first picture.) Here's what it looks like on Google Maps.
View Larger Map
Here's what my app collected
If you look closely you can see the individual data points. I get 10 samples per second so about one every 6 feet as I drove through this turn.