GPS flight recorders were used in the PPG, PL1 & PL2  championships as primary evidence.  431 tracks were recorded on seventy five MLR SP 24XC flight recorders very generously loaned to the organizers by the French federation; FFPlUM.  The device itself does not record accuracy but it is quite easy to see from the recorded tracks the kind of accuracy it was getting. 

 

In the first instance, poor reception is characterized by ‘spikes’ in the GPS altitude record which are very temporary, often lasting no more than a fix or two.  The position (and hence ground speed) is unaffected.  In the table below ‘unambiguous’ tracks may have the occasional spike but there is no doubt about the flight.  As the number of spikes in a track increases it becomes more difficult to detect whether a pilot has landed during the flight and to discern precise takeoff and landing times but it is unlikely to affect the pilot’s score.

 

This is then followed by ‘flat spots’ in the altitude record which are longer periods of poor reception.  The position (and hence ground speed) are often unaffected, but in the case of very poor reception there may also be a ‘spike’ in the track.  This latter represents a ‘risk to score’.

 

Occasional reception is characterized by a flat spot in the altitude readout and a straight line joining the occasional fix.  These tracks represent a probable risk to score.

 

In the table below, tracks with ‘possible partial score loss’ does not mean a score was lost, it means that there were long periods of very poor reception but not necessarily over ‘scoring locations’ such as turn points or gates. 

 

There was one case of mechanical failure where the device could be knocked and it would turn off.  This happened to a PL2 on takeoff and the pilot did not notice to switch it on again.  In that case he scored little because although marshals recorded the timed part of the task, he did not collect any secondary photographic evidence.

 

	Task 1	Task 3	Task 5	Task 6	Task 8	Task 9	Overall
Unambiguous	64.86%	71.23%	75.32%	80.30%	65.71%	74.65%	71.93%
Small number of spikes	27.03%	21.92%	10.39%	10.61%	21.43%	18.31%	18.33%
Large no. spikes but no risk to score	4.05%	1.37%	9.09%	3.03%	7.14%	4.23%	4.87%
Possible partial score loss	1.35%	1.37%	3.90%	3.03%	2.86%	1.41%	2.32%
Useless / DNS / Did not switch on	2.70%	4.11%	1.30%	3.03%	2.86%	1.41%	2.55%
	Task 1	Task 3	Task 5	Task 6	Task 8	Task 9	Overall
No risk to score	95.95%	94.52%	94.81%	93.94%	94.29%	97.18%	95.13%
Risk to score / no score	4.05%	5.48%	5.19%	6.06%	5.71%	2.82%	4.87%

 

Conclusion

 

There was more than a 95% success rate. Some of the useless tracks were from people who in fact did not take off in a task so all we have is some manoevering on the takeoff airfield.  The overall conclusion must be that the success rate was probably much better than the combined failure rate of cameras plus marshals observations as used in the ‘traditional’ method of data collection.

 

Considering that pilots did not have any opportunity to practice with these devices before the championships (eg to select a good position on the machine where it would permanently have good reception) the use of flight recorders in this championships must be considered a huge success.