Analytics Tracking Code

Wednesday, September 9, 2009

Accuracy of a GPS in a Car

So I have been having fun the past two days driving around Canyon Meadows (here in Calgary Alberta) with my GPS collecting information to post to Open Street Map. Although I have used the GPS for most of the summer as I scooted across the country with my bike, I have not done much with it in terms of actual mapping.

Now that I am officially back in town, I figure that a little bit of mapping would be fun to do, after all there is hordes of information missing even from the community within which I live. So last night, one of my boarders and I hopped in the car and drove the alleys of NW Canyon Meadows. I was slightly disappointed to discover that when I plugged the data in JOSM, the track was jagged and in areas, clearly wrong. What could be the problem? I have used this before to map the town of Empress Alberta without much difficulty.

It occurred to me while lying in bed that maybe the problem had to do with where the GPS was stationed in the car. On our midnight saunter, the gps was held in my passengers hand somewhere around or below his lap. Could this cause the problem? Maybe the unit needs better line of site to the satellites, and even the moderate amount of flesh and metal in the way was interfering with this.

Perhaps this is bang on because I just got back from another little jaunt where I had the GPS sitting on the dash. The result is considerably better and dramatically smoother. So the moral of the story is that the best place for your GPS while in a car is somewhere near the dash or somewhere with more or less clear access to a large chunk of sky. I suppose the more sky, the better.

As a side note, the unit was reporting on average +- 3 m the second time around.

No comments:

Post a Comment