Incorrect Return Routing - Massachusetts, USA

[storageanarchy]

Although I have meticulously attempted to find/fix the map problem that causes this, I simply cannot figure out why:

Travelling FROM "176 South St, Hopkinton MA" to "21 Jacques Rd, Tyngsboro MA" gets me the shortest route (along Depot St. in Westford),

BUT

Travelling in the reverse direction, always takes me a much longer route (up to 10 minutes longer). Further, if I drive the shorter route TO Hopkinton, the navigation doesn't recalculate the correct route until I cross Route 40 on Dunstable Rd in Westford.

All other GPS units that I've used find this route as the shortest flawlessly...

I have checked for red arrows that should be green, one way streets that should be 2 way, etc. and I just cannot find the map error that is causing this...

[xteejx]

If you provide a permalink, that will help an editor in your area to help you out. Also, please use your country's forum for location specific problems OR add the place name, state and country to the thread title.
It helps us pick things out much easier

Sent from my GT-I9100 using Tapatalk 2

[storageanarchy]

OK, thanks, xteejx

I apologize, but I don't know how to create a perma-link for a route... I will try to research the forum more to better understand how to work with it.

Thank you!

[ArlenBystander]

Here's the Permalink wiki page. Selecting multiple segments can be done either via Ctrl/Cmd-click or by toggling the multi-select option (the 'm' key) within WME and left clicking each segment in the route.

[xteejx]

No need to select segments, just a permalink of the area.

Sent from my GT-I9100 using Tapatalk 2

[storageanarchy]

Thanks -

I did repost this in the USA forum (there is no Massachusett sub-forum).

After further research, I find that the on-line Navigation will pick the correct route if I start 2-3 houses down from mine, OR if I set the destination only about half-way down the main highway I have to travel. Moving the destination just 1/2 mile further down the road, and the route reverts to the 10-minute longer version. Thus, I suspect it's not a map problem, but a glitch (or feature) of the route selection software.

I'm not sure how best to report this issue to the developers, though.

[harling]

Permalink (of the area near the destination, where the routing variations occur).

I didn't see any obvious problems either (other than some reverse-connectivity, which I cleaned up as a precaution). Waze does give a preference to major roads over minor ones--but on the return trip, I would still expect it to direct you to Scribner Road (Primary street) rather than all the way out to Route 3.

It also could be that Waze does not have enough real-world speed data for the longer route to make an accurate assessment. Without GPS tracks, Waze will impute a default speed for those roads & junctions, which may be faster than actual speeds. If you follow that route a couple times, it will provide Waze with the information it needs. And if that is indeed the problem, it should stop recommending that route (unless traffic dictates otherwise at the given moment).

[storageanarchy]

Thanks for the feedback.

What is interesting, is that WAZE tells me the ETA for its preferred Route 3 route is (let's say) 9:05AM when I leave my home. As I drive my preferred route, it keeps trying to get me back to Route 3, extending the ETA to perhaps 9:12 AM. Then when I cross Route 40, it recalculates the route to my preferred, and immediately tells me that my ETA is 8:55AM...10 minutes earlier than the original "long" route.

This morning, I chose to stop for gas at the gas station on the corner of Scribner/Dunstable road and Route 40. As I pullled in, the plotted route was through Graniteville. But after sitting at the pump for a couple of minutes, WAZE suddenly told me that there was a quickler route - down the Depot st. path of my preferred route!

I'm wondering if WAZE doesn't give up on a route based on the number of segments or something, although that wouldn't explain why going TO Tyngsboro would pick the right route and FROM goes the long way. Nor does it explain why it will select the quicker way FROM Tyngsboro if I select a destination on I495 BEFORE the Boxboro exit, but drop back to the longer Route 3 path if I select a point of 495 just PAST the Boxboro exit. I've seen similar happen if I select my starting location as 37 Jacques Road instead of 21.

Makes no sense to me...it clearly knows my way is faster, but it just doesn't want me to go that way!

[harling]

Thanks for the feedback.

What is interesting, is that WAZE tells me the ETA for its preferred Route 3 route is (let's say) 9:05AM when I leave my home. As I drive my preferred route, it keeps trying to get me back to Route 3, extending the ETA to perhaps 9:12 AM. Then when I cross Route 40, it recalculates the route to my preferred, and immediately tells me that my ETA is 8:55AM...10 minutes earlier than the original "long" route.

That could be explained by Waze's road-type preference. The trip may be long enough that Waze prefers to get you on a highway (Route 3), than have you travel so far on Primary roads before reaching anything bigger. (That Waze considers the Primary road shortcut at the end of the trip but not at the beginning, suggests that its search heuristic might be different at the origin and destination ends of the trip.)

This morning, I chose to stop for gas at the gas station on the corner of Scribner/Dunstable road and Route 40. As I pullled in, the plotted route was through Graniteville. But after sitting at the pump for a couple of minutes, WAZE suddenly told me that there was a quickler route - down the Depot st. path of my preferred route!

I'm wondering if WAZE doesn't give up on a route based on the number of segments or something, although that wouldn't explain why going TO Tyngsboro would pick the right route and FROM goes the long way. Nor does it explain why it will select the quicker way FROM Tyngsboro if I select a destination on I495 BEFORE the Boxboro exit, but drop back to the longer Route 3 path if I select a point of 495 just PAST the Boxboro exit. I've seen similar happen if I select my starting location as 37 Jacques Road instead of 21.

The total length of the trip can affect the cutoff point for lesser road types. When it recalculates along the way, the overall trip is shorter, which may allow it to consider taking enough Primary roads to follow the better route.

[storageanarchy]

Thanks for the added insights...but I'd say the algorithms need to be tuned. I've used probably 6 or 7 different brands/types of GPS, and WAZE is the first and only not to recognize the actual shortest time route from Tyngsboro.

Of course, I don't need WAZE to know the fastest route for this trip, since I do it daily. But the deficiency causes me to wonder where else it is taking me the "long" way.

Bummer...crowdsourcing the traffic is a cool idea, but it defeats the purpose if I START OUT with a 10 minute delay.

Is there a forum where I can log this bug, and this complaint?