Peter Hodson talks about the accuracy of GPS

Incorporating GPS into Coxmate products was a big project for us last year. Even today, the technology continues to advance and there are a few enhancements for GPS happening now. The main objective is to correct for the refraction through the ionosphere. The ionosphere is a portion of the upper atmosphere distinguished because it is ionized by solar radiation and in the absence of selective availability, can be the greatest source of range and range-rate errors for GPS users. By correcting for the refraction through the ionosphere, the GPS will give a higher degree of absolute accuracy. GPS systems are introducing land based reference points. These are precisely known and used to determine a correction factor. The European version of this is known as SBAS.

However Coxmate is not greatly concerned with absolute accuracy. You aren’t going to get lost using a Coxmate SX with GPS! We need high relative accuracy over a relatively small area – you rowing down your river or lake.

We know from experience that the quality of GPS devices varies significantly. Coxmate has selected ones which perform well. We use no smoothing to condition the signal – avoid damping the responsiveness of the speed reading.

The bigger technological problem that we deal with is due to the boat speed changing through the stroke. If for example you were rating ~24 strokes per minute: The stroke period would be 2.5 seconds and the GPS gives an update once per second.

?GPS Reading Graph

The graph above shows the boat speed over 6 strokes – black line. During this period there are ~15 GPS updates. The Coxmate SX updates speed every stroke. From graph you can see how some strokes receive 2 and some 3 GPS readings. The horizontal red line is the average of the readings received during the stroke. For example the first reading is average value of GPS Readings 1 and 2. The second stroke is the average of 3, 4 and 5. It can be seen that the speed value updated every stoke is oscillating by around +/-5%. The size and frequency of this oscillation will vary depending on rate. If you rate 20 or 30 per minute, then each stroke would have exactly 3 or 2 GPS readings and there would be no oscillation.

The Coxmate SX, when in its default setting will have smoothing set to ‘Auto’, where it adjusts the number of strokes that speed is averaged over according to rating. This optimises requirements for fast response and smooth signal.

The SBAS and systems like it will undoubtedly become standard in the future and it will make it easier for poorer quality GPS chips as used in phones to improve accuracy and responsiveness. However it won’t fix the stroke speed oscillation.

Enhanced by Zemanta

2 thoughts on “Peter Hodson talks about the accuracy of GPS

  1. Brett Smith says:

    Hi Pete – Following on from discussion we have had over the last year – do you have any data on the reliability and validity of your GPS units (Standard error of measurement %) and if so is this data published. I have done a little bit of analysis but only with the SpiElite units (we put them in the boats at International regattas and compared the data)

    • Rebecca Caroe says:

      Apologies for delay. Have been tied up with our Nationals.

      Initially we looked at spec’s for GPS modules, but found that they didn’t seem to be too relevant to their performance, and resorted to testing. We tried ~half a dozen different GPS modules and found few of them performed well in the 10-20 Kph range.
      Our supplier has firmware to utilise the offset corrections for ionosphere refraction and we plan to test to see if it offers any benefit. Given that we’re not very interested in absolute accuracy it may not be of much help, unless it reduces the ‘flutter’.
      I’d be very interested to hear more on your tests.


Leave a Reply

Your email address will not be published. Required fields are marked *