leaderboard left
leaderboard right
content topleft content topright

Train Your Nuvi ETA

Sep
11
2008

Did you know that you can train your Garmin Nuvi to provide better estimated times of arrival (ETA)? You can! Yes… it is true, your Nuvi can actually learn from your driving and adapt its internal database to provide better estimates of your arrival time. Want to know how to activate it, or how it works? Follow along…

How Long?

So imagine this… you and a buddy both have a Garmin Nuvi 760. You both have the same application version installed, and you both have the same map version installed. You’ve also double checked that you both have the same routing preferences set.

You both generate a route using the airport POI as the starting address, and your hotel as the destination. Both devices generate the same route, and show the same distance. That is expected. But then you notice that your devices show different estimated travel times! How can that be?

This got our heads scratching a few months ago, so we solicited feedback in the forums. After hashing out the possibilities, we formed a hypothesis that it is possible (as crazy as it sounds) that the Nuvi is learning about your driving patterns and adapting the ETA based on your previous driving habits.

The Test

So we devised a test which we felt would eliminate all other variables we thought might be a factor… things like traffic receivers for example. We also “took one for the team” and performed factory resets on a number of devices. After those resets, and eliminating things like traffic receivers, we felt we had a good benchmark. Sure enough, all of the devices showed the same estimated travel times for specific routes.

Then we went about our normal driving with the devices, and after a couple of days went back and calculated the same route again. Sure enough, not only were the ETAs different from before, but they were also different from user to user.

The Proof

In my mind, this was plenty of proof that the Nuvi was indeed learning from driver behavior and adapting estimated times of arrival. If you put the device into simulation mode you could even see in some cases where the GPS was expecting you to be exceeding the posted speed limit, visible on the map display.

But I still wanted more. Garmin has since confirmed to us that yes, our theory was pretty much on the mark.

“We do learn from a driver’s habits and adjust the ETA accordingly. We do this by observing what speed a driver typically drives on each speed category of road and use that to compute ETA.”

Brilliant! Garmin baffles me sometimes — with features like this they are often too humble. Quite often people will ask if there are devices on the market that can learn from their driving habits, and I’ve always told them “no”. Yet here is a device I already have several of that already has the feature. There are no settings to turn on; this works out of the box.

Humble Pie

Often people are looking for the device to “learn” other things– like routes we’ve taken. But this is certainly a good start. Garmin seems to be that quiet kid in the back of the classroom that seldom volunteers an answer, but when they are called upon always has the correct answer.

In fact, we often joke about how the Nuvi of today is “dumbed down” from the Garmin auto models offered a few years ago that had many more features. While many features have been streamlined, in this case the joke is on us and the Nuvi is much smarter than we thought. :)

Do you IQ?

The natural question becomes “How does this compare to TomTom IQ Routes?” IQ Routes is based on all TomTom rivers that have opted in, so the sample size is much, much larger than the Garmin that is just listening to you. Also the TomTom method will get applied into route decision making, and can change the route picked based on the IQ data. I’ve yet to see an example that shows the user data will actually impact a route choice on the Garmin. However the Garmin is based on your driving habits rather than everyone else’s.

Much of the credit for this article goes to one of our amazing forum members, gatorguy. He brought the hypothesis to the table and helped gather data from other forum members. Community data rocks. :)

15 Responses


  1. Makes sense. I always marveled at how accurate my nuvi was in forecasting arrival times. Much better than other devices.

    Rich Owings - September 12th, 2008
  2. Indeed. And here is another observation. Whenever I drive with a particular Nuvi 760 I always complain that it is way too optimistic and that I never meet the ETA. Guess what? Someone else in the family uses that Nuvi more frequently than I do– and guess who drives faster? :)

    Tim - September 12th, 2008
  3. Excellent…wondering if the same holds true for old Nuvi’s as I have a 360.

    Peter - September 12th, 2008
  4. An observation of IQ routes on my 730. The IQ logic does learn your habits, but instead of applying them to the same road class, the TT applies it to that section of road. So not only are you relying on other users speed, but also your own for roads you frequent. I’ve noticed routes changing over time as it gathers more speed data on the local roads, ‘as I drive them.

    George - September 12th, 2008
  5. AFAIK, it does.

    gatorguy - September 12th, 2008
  6. George, that’s comes as a surprise. I understood IQ routes were tied to the maps. I hadn’t noticed my IQRoutes changing yet. AFAIK, your road speeds are logged and reported back to TomTom, but do not update your device routing until you buy a new map.

    gatorguy - September 12th, 2008
  7. That would come as a surprise to me as well.

    Tim - September 12th, 2008
  8. I see no reference for the quote from “Garmin”. I would be interested to know if this is really something they are acknowledging.

    Adam - October 1st, 2008
  9. Adam, the quote is in the middle of the article:

    “We do learn from a driver’s habits and adjust the ETA accordingly. We do this by observing what speed a driver typically drives on each speed category of road and use that to compute ETA.”

    Tim - October 1st, 2008
  10. The quote was provided directly to me which is why there is no external reference. If you want to follow up (since you are a Garmin employee) you can contact Jake Jacobson, Senior Media Relations Specialist.

    Tim - October 1st, 2008
  11. Thanks Tim, and I actually work for Tom Tom.

    Adam - October 6th, 2008
  12. Then why are all of your posts coming from Garmin’s headquarters?

    Tim - October 6th, 2008
  13. I am hacked into the Gibson of course!

    Adam - October 6th, 2008
  14. Makes sense! I just got a 255 and took an alternate route to work one day that I knew was longer (due to stop lights, etc.) but it was the “recommended route”. The actual arrival time was 6 minutes later then the GPS initially thought. The following day, they suggested a different (and quicker) route to take. I think it does learn about your driving characteristics but also the route conditions (slower or faster then expected).

    Laura - March 5th, 2009
  15. Will the Garmin 255W7 purchased off the shelf contain the estimated time of arrival (ETA) feature and the distance calculation feature, or do I have to purchase additional item(s)?

    Thank you

    Jerry - August 7th, 2010



content bottomleft content bottomright