What advantage does the tech package offer?

[zwede]

Just wondering - has the navigation map data ever been updated yet? Do we know for certain?

Yes, it has. A friend of mine had a house built in January this year. The whole area is currently being built, so the streets didn't even exist until fall '13. He warned me no nav will find it (other people have gotten lost), but the Tesla nav found it just fine. The satellite picture was current as well.

 

[scaesare]

Yes, it has. A friend of mine had a house built in January this year. The whole area is currently being built, so the streets didn't even exist until fall '13. He warned me no nav will find it (other people have gotten lost), but the Tesla nav found it just fine. The satellite picture was current as well.

Interesting. I think I've had some changes reflected around here as well, but I wasn't sure. I expect that the center console will get them relatively quickly, as it's an online map delivered via 3G connectivity... it's the turn-by-turn on the dash I've wondered about, as that's a Navigon system.

As Brianman mentioned earlier, nothing ever seems to have been noted in FW updates, but I wonder if this is a "constant" thing. It was noted in the "internal ethernet" thread that when a successful connection was made to the car, one of the directories that was published from the dash computer was that map directory... so I wonder how often it gets updates...

 

[bollar]

Yes, it has. A friend of mine had a house built in January this year. The whole area is currently being built, so the streets didn't even exist until fall '13. He warned me no nav will find it (other people have gotten lost), but the Tesla nav found it just fine. The satellite picture was current as well.

I believe there have been updates, but since you're in Plano, I'll note that the system is missing almost all of the DFW connector. For example, the !!4E to 121W flyover that have been open for more than a year.

 

[paulkva]

I can't say for certain if/when they were updated but there was a highway interchange that was opened a few months back. Initially the nav was confused about where I was and what I was doing but now it routes me properly through the interchange.

My .02

As for the tech package, I opted for it primarily for the nav. Having the maps/traffic data/cell service has been a great thing. I live in one of the worst areas for traffic congestion (Northern Virginia/DC Metro) and being able to check the traffic via the maps is awesome. Not to mention that I use it atleast twice a day.

It might have been updated, but the updates are far from perfect. I have relatives who live in a new-ish development in Silver Spring MD (Poplar Run); Google Maps has it up-to-date but the nav system doesn't know anything about it. The development has been around for multiple years. Similarly, I can't get nav to route me on MD 200 (Inter-County Connector) even while I'm driving on it.

Live traffic on the center screen is awesome, uncannily accurate for me most of the time, and has definitely saved me lots of time in aggregate.

 

[seanahan]

If I had paid more attention in Computational Theory classes way back when, I might even go so far as to suggest it's NP-hard, or NP-complete. But then I've mostly forgotten what those mean...

Not to get too off topic, but you can solve the problem with Dijkstra's algorithm - Wikipedia, the free encyclopedia in O(V^2), where V is number of intersections. I'm sure they are doing something more sophisticated with heuristics to remove highly unlikely paths, but in the end it will just be an enhanced version of Dijkstra.

Most likely, the actual data they have is noisy, and so the route seems optimal to the computer, but in the real world is not. For example, if they are missing speed limit data, they could send you down a 30mph when there's a longer route that's 60mph the whole way. Also, the country roads are likely not as well paved, so while the nominal speed limit might be high, in practice you are driving a lot slower.

 

[tga]

Yeah, I knew it was O(n^2), but the second half of that post was pretty boneheaded... I guess I shouldn't comment on stuff I haven't studied in a 2 decades (NP completeness) when I'm tired and have a couple of beers in me.

Not to get too off topic, but you can solve the problem with Dijkstra's algorithm - Wikipedia, the free encyclopedia in O(V^2), where V is number of intersections. I'm sure they are doing something more sophisticated with heuristics to remove highly unlikely paths, but in the end it will just be an enhanced version of Dijkstra.

Most likely, the actual data they have is noisy, and so the route seems optimal to the computer, but in the real world is not. For example, if they are missing speed limit data, they could send you down a 30mph when there's a longer route that's 60mph the whole way. Also, the country roads are likely not as well paved, so while the nominal speed limit might be high, in practice you are driving a lot slower.