The S2s deployed in Oxford are showing a significant (2.5% to 5%) “long tail” of high speeds (43+ mph, 70+ km/hr), even the ones on 20mph streets, some in locations with significant traffic calming. Because these are a relatively small fraction of measurements, it seems possible that they are (mostly) the result of some kind of systematic error.
See for example the measurements here 9000005087 and the Google Streetview shows the quite aggressive tables at either end of the segment - and (not shown on Streetview) there is now a modal filter (bollard) at the eastern end, so cars are have to turn to enter the segment: Google Maps
I’m not convinced 4% of cars are getting up to 40+ mph here.
This is interesting. We can certainly look at the data for some of these streets, if you or others can give us more local knowledge to help evaluate the specific segments.
I have looked at many other devices in Oxford and none of these show this discrepancy, so it could be something to do with the set-up of the device on Howard Street which could be causing this.
In the past there were some issues with speed measurements arising from our night data, and we’ve recently started to purge some of the misleading data while we optimise the night count algorithms. I think we should already have done this for these devices, but we can take a look.
These five are deployed in the same area, in streets with similar characteristics, but three are showing a “tail” of 43+ mph speeds and two aren’t. One thought I had was that cargo trikes (we have a fair bit of cycle freight) might be detected as fast-moving car (because they are short compared to cars).
Telraam | James St
Telraam | husrt
Telraam | HOWARD STREET
Telraam | Bullingdon Road
Telraam | Magdalen Road - not much data yet
It is unlikely that cargo bikes would be classified as cars by our S2 devices - though it might have been possible with the older V1.
We’re still looking at the data
Thanks, will be interested to see if you can find anything. On an unrelated note, my first post in this thread has been marked as spam by some automated system… I don’t suppose you can look at that and see what’s gone wrong?
Sorry @DannyYee - there was a moderation issue because of the number of links, but I have tweaked that setting.
I needed to amend the post, but it should all be back now.
Thanks for posting and we will keep monitoring this issue