Placing nodes at speed limit changes

by **TonyG-UK** » Tue Sep 11, 2012 10:19 am

Has the concept of placing nodes on roads where there are (major) speed limit changes ever been floated and discussed?

The idea would be to allow Waze to more accurately form average segment speeds, increasing the accuracy of slow-down detection. For example, there are places where the speed limit changes from 60mph to 30mph within a quite lengthy segment. This is likely to have the effect of artificially lowering the average speed of the segment, making it less likely that Waze will automatically detect a slow-down.

by **richardjeeves** » Tue Sep 11, 2012 10:47 am

This exact thing happened to me today. I was going down a 60mph road about to pass through a small village (30mph) and I had a nice yellow line already there saying 34mph.

by **AlanOfTheBerg** » Tue Sep 11, 2012 12:45 pm

It has been discussed before. I believe in more than one thread and was not adopted because over the full length of the segment, regardless of speed change, the average travel time will still be the same. Some people would like it because in some situations as pointed out already, the speed limit change may sometimes be made obvious with a false slowdown notification.

The "more granular" speed data argument doesn't make sense because if it is a single long segment without a way to get off the road to take a different route, theres no need to know about a slowdown because you can't do anything about it. If the slowdown is significant then the entire segment will be flagged as such and Waze may choose to route you around before you get there. But once you're on it, you're stuck anyway so granular reporting of a slowdown 2km ahead doesn't make any difference.
--
If it looks strange, or I used bad grammar or misspelled anything, it's my phone's fault.

by **iainhouse** » Tue Sep 11, 2012 1:25 pm

Another problem is that another editor may come along, not realise why the road has an unexplained node in it and delete it.

by **TonyG-UK** » Tue Sep 11, 2012 1:41 pm

iainhouse wrote:Another problem is that another editor may come along, not realise why the road has an unexplained node in it and delete it.

That's why we have locks

by **harling** » Tue Sep 11, 2012 2:41 pm

AlanOfTheBerg wrote:[O]ver the full length of the segment, regardless of speed change, the average travel time will still be the same. Some people would like it because in some situations as pointed out already, the speed limit change may sometimes be made obvious with a false slowdown notification.

The "more granular" speed data argument doesn't make sense because if it is a single long segment without a way to get off the road to take a different route, theres no need to know about a slowdown because you can't do anything about it. If the slowdown is significant then the entire segment will be flagged as such and Waze may choose to route you around before you get there. But once you're on it, you're stuck anyway so granular reporting of a slowdown 2km ahead doesn't make any difference.

I could see an AM doing this for a chronic trouble spot, to prevent Waze from perpetually reporting traffic (and perhaps routing around it) due to an abrupt speed limit drop near the end of a segment. But this practice should be done advisedly--in this one location, to address this one specific problem--not habitually.

TonyG-UK wrote:
iainhouse wrote:Another problem is that another editor may come along, not realise why the road has an unexplained node in it and delete it.
That's why we have locks

And why it would be really nice to be able to leave notes on the map for other AMs, as pushpins and/or attached to map features such as segments or junctions.

by **AlanOfTheBerg** » Tue Sep 11, 2012 3:45 pm

harling wrote:I could see an AM doing this for a chronic trouble spot, to prevent Waze from perpetually reporting traffic (and perhaps routing around it) due to an abrupt speed limit drop near the end of a segment. But this practice should be done advisedly--in this one location, to address this one specific problem--not habitually.

It would never do this because of the speed average for that time of day over the entire segment. Having another junction in the middle does nothing helpful.

by **harling** » Tue Sep 11, 2012 9:05 pm

AlanOfTheBerg wrote:
harling wrote:I could see an AM doing this for a chronic trouble spot, to prevent Waze from perpetually reporting traffic (and perhaps routing around it) due to an abrupt speed limit drop near the end of a segment. But this practice should be done advisedly--in this one location, to address this one specific problem--not habitually.

It would never do this because of the speed average for that time of day over the entire segment. Having another junction in the middle does nothing helpful.

Suppose you have a two-mile-long segment at the end of a major highway, which terminates at a roundabout. The speed limit for the first 1.9 miles is 60mph, but it drops to 30mph for the last 0.1 mile before the roundabout. If everyone were to drive the speed limit along this single segment, the average speed for that segment would be 57mph.

However, each time a car traverses that segment, it spends the last 500ft. at 30mph--significantly below the average speed. Next thing you know: *ding!* "Are you in traffic [Yes, Definitely, or Yes] ?"

Adding a junction where the speed limit changes would segregate the two very different average speeds, and allow Waze to tolerate the slow-down. (The last 0.1 mile segment would have an average speed of 30mph, and no traffic alert would be generated.)

by **AlanOfTheBerg** » Tue Sep 11, 2012 9:19 pm

harling wrote:Suppose you have a two-mile-long segment at the end of a major highway, which terminates at a roundabout. The speed limit for the first 1.9 miles is 60mph, but it drops to 30mph for the last 0.1 mile before the roundabout. If everyone were to drive the speed limit along this single segment, the average speed for that segment would be 57mph.

However, each time a car traverses that segment, it spends the last 500ft. at 30mph--significantly below the average speed. Next thing you know: *ding!* "Are you in traffic [Yes, Definitely, or Yes] ?"

I disagree with your assessment of how Waze will mark the segment. Because you still traverse the entire segment within the average speed tolerance average of 57mph, it won't pop up like this. I haven't ever seen this happen for streets which drop from 45 to 20 for school zones in the middle of segments, so I don't see why it would do it in the case you've presented.

by **harling** » Tue Sep 11, 2012 9:44 pm

AlanOfTheBerg wrote:I disagree with your assessment of how Waze will mark the segment. Because you still traverse the entire segment within the average speed tolerance average of 57mph, it won't pop up like this. I haven't ever seen this happen for streets which drop from 45 to 20 for school zones in the middle of segments, so I don't see why it would do it in the case you've presented.

IIRC, the client disregards the slowest 50 (or was it 30?) meters of travel on a segment, to prevent it from interpreting stop signs and lights as traffic. So unless the school zone is that long, it would barely affect the calculation. (That's one of the reasons my example uses longer distances.)

If the Waze client only evaluates your average speed after having traversed a segment, then you are correct: your average speed over the entire segment will be identical to the speed recorded for the segment, and no traffic will be detected.

But if you have ever received that *ding* "Are you in traffic?" pop-up while in the middle of a segment--before your average speed for the entire segment can possibly be calculated--then it seems to me that the client must be evaluating vehicle speed with finer granularity. Perhaps it triggers the moment your vehicle has traveled below some threshold percentage of average speed for more than 50m? If it works anything like that, the scenario I described would indeed trigger a report for every Wazer.