Speed measurements with S2 - generally to high?

We measured with a S2 device for a week the speed and compare the v85 values with those of commercial dialog display counters; 2 of the latter were covering the same street section witch has a speed limitation of 30 kmh; one placed about 50m, the other 250m away of the place where the Telraam S2 is counting. The standard deviations of both the commercial devices and Telraam are quit low, but Telraam values are consistently 30% higher than the measurements of the commercial devices. This is very disappointing. We had another device in a small side street, and also here have the (subjective) feeling that the speeds measured by S2 are too high.
The device was installed 2 weeks earlier; Segment is: 9000005808 A.

Enclosed is the table of the data and computation (pdf; excel not allowed?)
Table comparision.pdf (125.0 KB)
.

Interesting data. I have not experienced any like this. For us te margin seems to be well in between the 10% error - on all devices.

1 Like

Hi Alf, those are remarkable findings and disappointing indeed. I’m curious to know what the other device uses in terms of measurement technique, but I’m assuming it must be LIDAR, which is definitely reliable for speed measurements.

The fact that the deviation is consistently +/-30% points to a systematic error. I can’t see anything bad for this location, explaining why speed would be off. In fact, if we see errors, we actually expect Telraam to underestimate the speed.

The only thing I can imagine that could explain the error, is both sensors aren’t measuring the same objects (?). With Telraam, we only do speed for cars, so no trucks. The other sensor may classify light commercial vehicles as cars or Telraam may be classifying some bigger cars too quickly as a large vehicle. This is important for 2 reasons: LCV are typically slower and, secondly, the average length of what’s identified as a car by Telraam is lower. That parameters is relevant to calculate the speed.

Still, I don’t think the would explain the full difference; maybe a bit, but not all. What also is somewhat odd is that on the days Telraam has a lower V85, the other device has a higher V85 (compare 29/10 with 23/10)…

Puzzling…

1 Like