Why GPS without inertial navigation fails.
I am new to the boards and I didn't do a search so if I am repeating something that has already been said then too bad! :)
Anyhow, I have seen a few posts here questioning why their car appears off the road and/or "behind" your actual position. Most times this is not due to software. Commercial GPS is inharently inaccurate. Years ago position data was intentionally false so that private users--or other governments--could not utilize governments new toy against them. The Government has since inacted a bill that removed that false position information; however, commercial applications are only offered a single datastream from satellites and are restricted by the laws of physics for the data rate and any obstructions blocking your receiver from the satellites.
Government codes allow for 2 data streams,--commercial and coded--which significantly increases the accuracy of the position data yet for their needs is still unsufficient. The solution is to take Inertial navigation--gyros are very accurate relative to a specific position--and blend it with GPS. With this system you can get down to very specific positions.
This accuracy of this system is magnitudes better in poor (and good) whether and when the ionosphere is at its peak width around 1-3pm--ever notice your positions get worse around this time? The increased accuracy in this system is caused by the fact that the blended solution is still heavily influenced by an accurate inertial system even though the GPS may be off 50-100 feet.
I guess what I am suggesting is that if you really want a good navigation system go look for something that blends inertial navigation with GPS. I am not sure if there are any commercial applications--something that would easily plug into your computer--but I have to believe they are available. I know infiniti uses a blended solution in their navigation.