Using Interstate Mile Markers to Calibrate an Odometer

[Mr. Kite]

I recently drove about 1,000 miles in my HiHy (from CO to AR) and used the handheld GPS to calibrate my odometer along the way. I would record the numbers from the trip odometers from the handheld GPS and the HiHy. These were all nonstop datasets consisting of just under 100 miles each (after 100 miles, I lost the digit after the decimal on the handheld). For each of these sets, I plotted GPS odo vs. HiHy odo. I did not include 0,0 as a data point since there is some error in my manually resetting the two devices. I then obtained the slope from a straight line fit to get the error. Through 5 little sets of data, the slopes were 1.0344, 1.0351, 1.0337, 1.0329, and 1.0351. This averages to 1.0342 (the HiHy's odometer needs to be increased by 3.42% to be accurate). I will collect more data on my return trip before I update my tank entries.

[fuzzy]

I'd like to add a few comments to this ancient string. As an engineer (not civil) and former road surveyor (college summer job slope-staking dirt logging roads), I feel fully unqualified to speak to this issue.

- I can't answer whether or not vertical distance is included, but the impact should be very small: 0.12% on a constant 5% grade, and 0.24% on a 7% grade. Interstate highways should never be steeper than 7%, and such steep slopes should account for only a small portion of your comparison distance;

- On my past cars, the odometer and speedometer have had significantly different error factors. The two current cars are much better;

- Most of the mile markers I have used have been very consistent. A few seem to have been slightly misplaced, usually due to local terrain or obstructions, but many more are simply absent or not visible;

- When roads are rebuilt and moved, the length is often changed, usually shortened. This can be significant in winding canyons. Look for oddly placed markers with markings such as "MP 107.57 Ahead, 107.41 Back". The difference is the amount the road was shortened. Unfortunately, most of these "busts" are unmarked, or not prominent enough to catch the attention of drivers with more important things to watch;

- On divided hiways, the two directions can have significantly different lengths, especially when the two sides are placed thousands of feet apart on opposite sides of canyons. A driver has no idea which side, if either, has the accurate markers. For calibration, stick to relatively straight or parallel stretches or to non-divided roads.

I suggest using long stretches of road, but check many intermediate points and throw out any segments that appear to be "busted".

--Dean

[Aether glider]

Wouldn't use a GPS for this purpose. Way to many chances for errors. PDOP, terrain, obstructions, etc.

Unless you got a Trimble or other high end gps.

[Mr. Kite]

Originally Posted by Aether glider

Wouldn't use a GPS for this purpose. Way to many chances for errors. PDOP, terrain, obstructions, etc.

Unless you got a Trimble or other high end gps.

What would you use? I collected my data starting in eastern Colorado after it leveled out and throughout Kansas. I have found documents online that instructed dealerships to do the same to calibrate odometers that were off. Collecting lots of data points minimizes error. I'm confident in my numbers.