The first “real-world” test of the GPS Tracker was this last weekend when we went for a bit of a cruise through the state forest on our way home from Toodyay. It was a good test of how it performed recording slow, winding tracks as well as extended periods without GSM  coverage.

The report generated from the trip can be found at:

http://telemetry.tribalchicken.net/report.php?tid=3765

As can be seen from the report (Which is actually 2x seperate trips merged together – we stopped for a while at one point) the device works rather well at recording the general attribues – time, speed, etc. and the map looks great when you’re looking at a zoomed out, general overview:

When you get a bit closer, however, the precision leaves a bit to be desired. This is accentuated by the fact that it only records every 30 seconds, and the map simply joins the dots leaving you with an interesting jagged line:

Another problem which I had known about (From the way I had written the code) but never had the chance to test properly, is when it builds up a large buffer of pending events to send to the server like it did when we had no reception for 1.5 hours. It spent 10 minutes of the drive home firing off the buffer to the server, but not recording any new events:

As can also be seen here in the database:

This is great because it means the buffer on the SD card works, but bad because I need to  change the program to still record events while sending (Perhaps using interrupts – will investigate this).

Apart from those problems (Which I kindof already knew anyway) I can’t really complain about much.

I may look into relocating the external temperature sensor though, as currently it’s located underneath the bulbar. It turns out then the sensor easily gets wet and records lower temperatures if you go through a puddle

All in all I think it works well – will just need to investigate interrupting the transmission routine and perhaps play around with shortening the interval between events.