For those interested in the MO of how KA72 was involved in calculating the speeds of the sailors at the KA Sail Caloundra Wind & Surf Speed Challenge held this last weekend at Burrum Heads in Queensland, here is the lowdown.

I needed to come up with an offline version of the KA72 speed reader that would work on a laptop on location to process and give results for all the GPS sailors (about 33 this year.)

We also wanted to make it as painless as possible for sailors to get their data analysed. No standing in lines or mucking about!

So, after day one of the competition, those sailors that wanted to gave me their GPS, which was put into a ziplock bag with their name on it, and I took these all to my secret analysis location, and ran them through. This involved removing the cards from the Navis and copying the files locally for processing, or using GPSResults to download files from Garmins via serial cable.

As each file was transferred to computer, it was logged to the sailor's name in my KA72 Comp client, and the results were calculated according to the event organiser's specifications. In this case, we were asked for a 2 second peak, 5x10 average, nautical mile, and one-hour speeds, all to be calculated on trackpoints readings. (No doppler in this year's event.)

The actual calculations are guaranteed to be identical for each file, no settings to tweak, etc. Even so, there were some odd results that needed to be checked manually. One sailor recorded a 2 second peak of 61 knots in the speed reader. When I opened it in GPSResults, the reported 2 second peak was over 100 knots! So I was happy about that. There were some serious trackpoint errors in that file that no software was going to detect, so there is always a need to verify results, especially prize-winning ones.

The KA72 comp client allows me to override the calculated results for these boundary conditions, and the final tally contains about half a dozen corrections (across all divisions) though not every file has been completely validated.

All in all, the track analysis process took only about half an hour, or about one minute per GPS (taking out card, loading, analysing, recording times.)

Soon, the complete set of files relating to this event will be available for download. In the meantime, if there is a particular file you are interested in please feel free to contact me.

If you have any questions about the analysis process, please ask in this thread so everyone else can see the question and answer.

Dylan Tusler.