I Have Made a Routing Discovery

 

I've made a discovery in the apparent way routes are calculated. Many of you have complained that various Garmins will refuse to route a road even when it is decidedly shorter and faster. Even with Trends on and repeatedly using this road.

Before I make any statements my hummingbird hiney can't back up, I'd like for folks to make some observation for me. Really only one. Does your pet peeve road show a Speed Limit on the screen, or is it one of those roads, that while having a posted speed, doesn't show on the screen?

Also, for those units where you can edit posted Speed Limits, try adding a speed limit for the road and see if this makes any difference in the routing choices.

Please followup to this thread.

Thanks.

--
Frank DriveSmart55 37.322760, -79.511267

info

thanks for the info,PHRANC

--
I may have my user name and password confused or switched! Can I get confimation of my user name????

You could be on to something

I had to take a different route back from our trailer, due to road construction, this past summer. I used Trip Planner to force the route and I was surprised at how long it said it would take to get home. I did not have any time set for duration at each point.

The route took me through one or two "Blink and you'll miss it" hamlets and the ETA dropped by about 6 minutes. In reality, it took me 2 minutes tops and likely less than that, to get through these places.

I don't believe that I had speed limit for these roads at the time, so I can't comment. However, I expect that either the speed, that Garmin had was terribly wrong, or they Garmin's distance through "Blink and you'll miss it" was way to long.

Good luck.

--
DriveSmart 65, NUVI2555LMT, (NUVI350 is Now Retired)

Look at trip plannet

GPSgeek wrote:

I had to take a different route back from our trailer, due to road construction, this past summer. I used Trip Planner to force the route and I was surprised at how long it said it would take to get home. I did not have any time set for duration at each point.

The route took me through one or two "Blink and you'll miss it" hamlets and the ETA dropped by about 6 minutes. In reality, it took me 2 minutes tops and likely less than that, to get through these places.

I don't believe that I had speed limit for these roads at the time, so I can't comment. However, I expect that either the speed, that Garmin had was terribly wrong, or they Garmin's distance through "Blink and you'll miss it" was way to long.

Good luck.

I do trip planner on Base camp and it automatically puts 30 minute stops for each place I have on the rout. See if this happened to you.

--
Mary, Nuvi 2450, Garmin Viago, Honda Navigation, Nuvi 750 (gave to son)

n /a

I'm a frequent complainer about my Garmin routing me past a road that is both shorter and faster to a longer route with more traffic lights and more traffic. I have recently found that, in this case, it will take the road that I want if I'm approaching it from the other direction when I tell the GPS receiver to find a route home (coming from the road that I want to avoid).

That seems to tell me that the GPS is willing to route on the roads that I want, it just has a strong preference for the longer, busier, more traffic light route. At least it doesn't try to get me to make a U-turn to go to the undesired route.

Unfortunately, I can't answer the question that you ask, the GPS in question is a 250 and does not have a speed limit display. Although it does seem to know the speed limit on the route that I want, whenever it is routing me the long way and I turn on to the route that I want it does the "recalculating" thing and the new ETA is a couple of minutes less than the old ETA was, and the ETA reflects the limit on that road. (The mileage drops too, and that should be the determining factor, since this occurs when I have the nuvi set to find the shortest route, not the fastest (but it makes the same bad choice if set to fastest)).

Same here

Frovingslosh wrote:

I'm a frequent complainer about my Garmin routing me past a road that is both shorter and faster to a longer route with more traffic lights and more traffic. I have recently found that, in this case, it will take the road that I want if I'm approaching it from the other direction when I tell the GPS receiver to find a route home (coming from the road that I want to avoid).

That seems to tell me that the GPS is willing to route on the roads that I want, it just has a strong preference for the longer, busier, more traffic light route. At least it doesn't try to get me to make a U-turn to go to the undesired route.

Unfortunately, I can't answer the question that you ask, the GPS in question is a 250 and does not have a speed limit display. Although it does seem to know the speed limit on the route that I want, whenever it is routing me the long way and I turn on to the route that I want it does the "recalculating" thing and the new ETA is a couple of minutes less than the old ETA was, and the ETA reflects the limit on that road. (The mileage drops too, and that should be the determining factor, since this occurs when I have the nuvi set to find the shortest route, not the fastest (but it makes the same bad choice if set to fastest)).

I find the similar thing here. However, no matter which direction I come from, it refuses to route on this road until you actually get on the road.

--
Frank DriveSmart55 37.322760, -79.511267

Not Related

mgarledge wrote:
GPSgeek wrote:

I had to take a different route back from our trailer, due to road construction, this past summer. I used Trip Planner to force the route and I was surprised at how long it said it would take to get home. I did not have any time set for duration at each point.

The route took me through one or two "Blink and you'll miss it" hamlets and the ETA dropped by about 6 minutes. In reality, it took me 2 minutes tops and likely less than that, to get through these places.

I don't believe that I had speed limit for these roads at the time, so I can't comment. However, I expect that either the speed, that Garmin had was terribly wrong, or they Garmin's distance through "Blink and you'll miss it" was way to long.

Good luck.

I do trip planner on Base camp and it automatically puts 30 minute stops for each place I have on the rout. See if this happened to you.

This is nothing to with the problem I described. The units don't insert stop times, unless you use the built in Trip Planner and manually enter stop times. The things GPSGeek is seeing is because, not knowing the actual speed limit, the unit takes a best guess at the speed limit based on the road type.

--
Frank DriveSmart55 37.322760, -79.511267

didn't use basecamp

mgarledge wrote:
GPSgeek wrote:

I had to take a different route back from our trailer, due to road construction, this past summer. I used Trip Planner to force the route and I was surprised at how long it said it would take to get home. I did not have any time set for duration at each point.

The route took me through one or two "Blink and you'll miss it" hamlets and the ETA dropped by about 6 minutes. In reality, it took me 2 minutes tops and likely less than that, to get through these places.

I don't believe that I had speed limit for these roads at the time, so I can't comment. However, I expect that either the speed, that Garmin had was terribly wrong, or they Garmin's distance through "Blink and you'll miss it" was way to long.

Good luck.

I do trip planner on Base camp and it automatically puts 30 minute stops for each place I have on the rout. See if this happened to you.

I don't use Basecamp Trip Planner. Depending on the trip, I prefer to either save points on the Nuvi or create a dedicated trip POI. I then use these in Trip Planner on the Nuvi. I'm only using Trip Planner to coordinate the trip, rather than plan it.

Trip Planner in Basecamp is useless compared to Streets & Trips planner.

--
DriveSmart 65, NUVI2555LMT, (NUVI350 is Now Retired)

average rate of advance, not speed limit

phranc wrote:

... not knowing the actual speed limit, the unit takes a best guess at the speed limit based on the road type.

While lots of folks posting on this forum pass around the assumption that the Garmin units base route selection and ETA estimation on the database speed limit numbers, I think it is not true, and have never seen credible evidence of it.

As for evidence that it uses something else: consider that many, many road segments have no speed limit coded in the data base at all--yet somehow it uses them. Further, if you go into the simulation mode, and watch the speed your simulated drive is taking you at, it won't generally be at the speed limit. Further, I've noticed on the road between here and Santa Fe that the ETA-generating speed assumption differs on two stretches of I-25 which have exactly the same speed limit--proof that it is relying on something other than the speed limit in the data base (as for all the segments of I-25 in question it in fact has speed limit in the data base and displays it).

I've called this alternate number I infer to be in the database "average rate of advance" not because I know that is what Navteq and Garmin call it, but for ease and clarity of discussion. Think for a moment--on an urban road interrupted by frequent stops, direct use of speed limit for ETA estimation would be fatally flawed--what you want is a best guess of average rate of advance.

If anyone posting here has evidence they can cite that the unit --ever-- uses speed limit in a routing or ETA calculation, please post it. I'd be happy to be proven wrong.

--
personal GPS user since 1992

I too

I too would like to see evidence of the Nuvi using speed limit data but alas I think it's calculating ETA based on road type coding as it has for many years now, and long before speed limit data was ever on the device.

--
Nuvi 350, 760, 1695LM, 3790LMT, 2460LMT, 3597LMTHD, DriveLuxe 50LMTHD, DriveSmart 61, Garmin Drive 52, Garmin Backup Camera 40 and TomTom XXL540s.

evidence to the contrary

t923347 wrote:

...but alas I think it's calculating ETA based on road type coding as it has for many years now, and long before speed limit data was ever on the device.

I believe that ETA is based on the speed limits. While perhaps it is common practice to just post based on what one wants to believe, let me explain why I believe this. I have found that, on Interstate highways, if I set my cruise control at the posted speed limit and drive for a significant time at the posted speed, the ETA does not change. This is the case no matter if I'm driving at a posted speed of 60 or at a posted speed of 75, or any other posted speed. When I cross into a new state with a different posted speed, I have to match the new posted speed to keep the ETA consistent. If I speed up or slow down the ETA will slowly change. And this is true not just for different states that set different maximum Internet speeds, but also when the speed limit changes while in the same state.

And, of course, if I have to slow down due to the speed limit being reduced in a work zone, (a temporary condition) I see the ETA creep up.

I don't see any way that the nuvi could behave this way other than using the built in speed limits. It seem too much of a logical stretch to think that the nuvi would have some strange algorithm that seems to know when the speed limit changes on the same type of road (Interstate) and is able to compute a correct ETA based on that algorithm when they could just use the speed limits that are now in the data. And, if they are using speed limits when computing the route on Interstates, then I expect they are doing the same for all roads that they have speed limit data for.

I Agree

I won't requote all of that, Frovingslosh, but there is one note I would like to add. Several years ago, back when the Nuvi 760s and 765s came out, it was discovered that they "learned" driving habits, including average speeds the drivers drove. This was also confirmed by Garmin, that the units "learned" habits.

I can duplicate the same results you got, after my unit's "learned" that I tend to drive ~5-8 mph over the limits. ETAs would be consistently shorter than ETAs calculated over the same route by Delorme SA.

Another example is where I used the unit I owned in my school bus for a year. In VA school buses are limited to 45 mph, (or 60 when the speed limit was 60 or higher). I drove predominately at the 45 mph, since I took few trips.

After using it on the bus daily for quite some time, I noticed when using it for trips in my personal vehicles that ETAs were consistently way too long for the trips. It remembered that I always averaged 45 mph or less, even on roads posted higher and calculated accordingly.

--
Frank DriveSmart55 37.322760, -79.511267

True

phranc wrote:

I won't requote all of that, Frovingslosh, but there is one note I would like to add. Several years ago, back when the Nuvi 760s and 765s came out, it was discovered that they "learned" driving habits, including average speeds the drivers drove. This was also confirmed by Garmin, that the units "learned" habits...

As I remember, member DorkusNimrod has stated this often in the past as well and he adds the reminder that a hard reset of the nuvi will lose these learned habits.

Driving habits

Several bits of info.

(1)What is the myTrends feature and how does it work?

01/15/2014
The myTrends feature is available on select Garmin products. When you travel regularly to saved destinations, your device will, over time, begin to figure out where you’re going even without your telling it. Your device will provide a predicted route which will display in the information bar at the top of the map screen. myTrends provides time of arrival and relevant traffic information in the information bar.

Please note: myTrends is a feature that does not work out of the box when you first begin using your device. In order to effectively use the myTrends feature you must meet the following requirements before myTrends will display arrival and relevant traffic information:

Your frequent destinations must be saved in Favorites or Saved on your device, i.e. Home, Work, etc.
A regular pattern of driving habits must be established before myTrends will display on your device:
Daily regular pattern of driving takes at least three days of usage
Weekly regular pattern of driving such as a different destination on different days would take 2-3 weeks of regular usage
Example of myTrends:.

Press the information bar at the top of the map screen in order to display alternative routes if one is available. Here you can press Go! to activate a route to your myTrends destination:

Once you press Go! the device will begin providing turn by turn directions to your destination.

(2)What is the timeframe for a myTrends event to display on my device?

After you have an established pattern of driving habits your device will display a myTrend event within a two hour window. As an example if you leave work for home at 5pm each day, the myTrend event will display on your device as early as 4pm and as late at 6pm.

(3)Can I train my Garmin to provide better estimated times of arrival (ETA)?

04/07/2014
Your device can actually learn from your driving and adapt its internal database to provide better estimates of your arrival time. The device will learn from a driver’s habits and adjust the ETA accordingly. This is done by observing what speed a driver typically drives on each speed category of road and using that data to compute ETA.

--
Charlie. Nuvi 265 WT and Nuvi 2597 LMT. MapFactor Navigator - Offline Maps & GPS.

category, not limit

charlesd45 wrote:

...The device will learn from a driver’s habits and adjust the ETA accordingly. This is done by observing what speed a driver typically drives on each speed category of road and using that data to compute ETA.

Regarding our discussion on the use or lack thereof of speed limit coding, I'll point out that the quoted material (I presume this is direct from a Garmin source) speaks of "speed category", not of "speed limit".

I'll also point out that my own personal ETA-creep observation, on the Interstate between Albuquerque and Santa Fe, detected a change in the creep rate between two segments of I-25 which had the exact same posted speed limit, had the same database coded speed limit as displayed on my Nuvi, and had me travelling the same speed with cruise control engaged on my 2002 Audi A4--which has a cruise control which pretty quickly compensates out hill or wind-induced variation (from behavior I think it has PID characteristics, which were clearly lacking in the cruise control on older cars I've driven).

ETA-creep observations reporting earlier in the thread are inherently confounded with possible variations in the database speed category (hint: it is clearly not just "interstate"). My observation, on the other hand shows clear variation in the face of a fixed speed limit, and lacks that form of confounding.

--
personal GPS user since 1992

I also watch ETA creep.

I also watch ETA creep. However, I don’t put much faith in highway ETA creep. When comparing driving at 65mph versus 70mph, you only get 1 minute of ETA change after about 13 minutes of driving. Under those conditions, minor inconsistencies in speed can have big effects in ETA creep.

The same is not true in towns. Every red light seems to directly translate into an increase in ETA with never a reduction in ETA. That makes me believe their algorithm fails to account for average transit time through lights. I believe that also explains the unjustified preference for street routing when semi-adjacent highway routing is obviously faster.

Great info!

Great info!

red lights

zeaflal wrote:

Every red light seems to directly translate into an increase in ETA with never a reduction in ETA.

Yes, I have also seen that Garmin completely ignores red lights in ETA calculation. Unfortunately, it's routing seems to favor the major roads that are more likely to have red lights on them than alternate roads (in my experience) no matter if you have selected "faster" or "shorter". I have friends who tell me that Tomtom is much better about this and seems to have a average delay time built into each red light.

And yes, the ETA changes quite slowly when the difference in speed is around 5 miles/hour. But it is certainly measurable. I didn't come out and say it, but my reason for using the Interstates in this example is that these are roads where I can set cruse control and run at a set speed for times long enough to make these observations, without slowing down or stopping for red lights, speed traps, or similar driving events that make this harder to measure on other roads.

model dependent

zeaflal wrote:

Every red light seems to directly translate into an increase in ETA with never a reduction in ETA. That makes me believe their algorithm fails to account for average transit time through lights.

While the database largely obtained from Navteq is obviously the same, there are substantial ETA-estimate differences between Garmin models in my small range of experience.

With my older models (mostly a long-used GPS V, but also, I think, my 855) I saw in-town behavior in Albuquerque somewhat as you described, with the average bias on street-level ETA error to be on the low side.

However, with my 3790LMT, at least after a few months of "learning", the around-town average ETA error is about zero, and I see cases of reduction in ETA all the time (about as often as increases).

On the other hand, apparently I've not driven enough highway miles, or the learning got over-impressed with just how slow I drive, as on interstate travel the error is generally to give too high an ETA.

Just in case anyone doubts this model does some sort of learning, I'll mention that my wife and I have the same model, with the same firmware and maps, and that hers will give a different ETA than mine for the same destination (generally sooner on hers...)

--
personal GPS user since 1992

My Speed Higher Than Posted

phranc wrote:

I can duplicate the same results you got, after my unit's "learned" that I tend to drive ~5-8 mph over the limits.

I tend to drive about 10-12 MPH over the limit, when I'm on home turf in Ontario and when conditions permit.. This both on major highways and rural roads. The police here will generally ignore that overage. I will slow down considerably through the city, villages and hamlets, for safety reasons and out of courtesy to the residents, especially in small communities.

In the US, I lock the cruise at no more than 2 MPH over the limit. Enforcement is often tighter and there's always the old "Out of State License Plate Targeting" thingy.

Despite that, my ETAs are higher than they should be.

Has anyone ever gone to Garmin and tried to get an answer?

--
DriveSmart 65, NUVI2555LMT, (NUVI350 is Now Retired)

Someone did ask Garmin