Waze suggesting a LH + long drive instead of simple RH turn

Map Editors Community Australia New South Wales
1

Hi,

The intersection is here: https://www.waze.com/editor/?env=row&lon=151.11100&lat=-33.93607&layers=3045&zoom=8&mapProblemFilter=0&mapUpdateRequestFilter=0&venueFilter=0&segments=308480794&mapUpdateRequest=6384852

There are two reports, one’s mine but I suspect the other is similar to mine, only 4 hours later. I was heading towards Bexley North - all it takes is a R/H turn onto Bexley Rd.

Waze however was telling me to turn left, then turn into some streets, do a U-turn and return to Bexley Rd. I thought there might have been a Right Turn restriction at the intersection but everything looks correct.

I’m stumped.

2

Hi, The route speed script shows quite varied segment times for right hand turns at this intersection on Saturdays (11 June being a Saturday). It can be quite low in the order of 42s and above 184s. I say above because that 184s produced a route that took the right turn, but in another result I got the same route as you did (i.e. a LH turn) so the right hand turn time must exceed 184s… (1230 on a Saturday is one period that gives that route.)

I’ve just come across a similarly odd route in the ACT where Waze wanted to avoid a RH turn. see https://www.waze.com/editor/?env=row&lon=149.09045&lat=-35.34951&zoom=4&mapUpdateRequest=6435676
A few weeks back I has similar issues with the right turn from Hindmarsh drive into Yamba drive just to the east of this UR. I was directed to go straight through turn left, right, left, u-turn and back to the intersection to turn left.

In summary, its the segment data that is doing this. Not any error in the map - so nothing for us editors to fix.

3

Hi AliG,

Thanks for your reply. I’ve installed the script now too but the suggested data doesn’t make much sense. There just isn’t that much traffic on that exit on the weekends unless the M5 tunnel is blocked with heavy traffic and people turn off to avoid it (it’s the last exit before the tunnel).

I wonder if the anomaly is perhaps caused by the rather unusual mapping of the intersection. It’s composed of a rather large number of short segments with what I assume would be less accurate speed data (GPS units aren’t that accurate) and this might contribute to a more complex routing problem.

I wonder if the routing wouldn’t be better if all five intersecting segments met at one point, after all, this is a single set of traffic lights.