Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Elon "About to end range anxiety"

This site may earn commission on affiliate links.
And things like "last successful charge".
Tesla really needs to either do their own database of chargers or integrate one of the big ones into the car.
Then all I need to click is in the setting "CHAdeMO adapter on board" and I'm good to go :)
Put a (passive) card in the CHAdeMO adapter that the car can ping. Then you wouldn't need to enter that info "manually".
 
Trying out the trip planner right now. It has promise. It's badly undermined by the inability to choose different routes to your destination.

I've already saved at least an hour and a half by avoiding the Interstate, and taking a more direct route. Otherwise I'd have to go a long way off my path to get to another Supercharger. The system was telling me I couldn't make it to Utica, so I disabled navigation and used my iPhone. Once I got across the US border it started cooperating more, but still tried to do a lengthy side-trip to the Interstate.

On the other hand, at my Utica stop (easily arrived with 30% left despite being told I can't make it!) it gave a nice reassuring message after some charging that I had enough charge to get to Albany. That message only came through on the touchscreen though... would have been nice if a notification came to my phone (thought it was supposed to do that?).

For the last leg, it has me going 20 minutes past my destination to get to another Supercharger, then backtracking to the hotel. How about making sure there's enough reserve to get to the Supercharger, but letting me check in first??? At least I can remove the extra stop from the trip.

Twice now I've had the nav system hang up, calculating route and never finishing. I had to abort navigation and reenter the destination.

Twice the system failed to give voice announcements for upcoming turns (bear left/bear right types). It's never done that before.

Despite that, there was one occasion where it told me to continue straight - I hate when GPS sytems do this. It's really saying, "In one mile, don't do anything." In this case there wasn't even a road to turn on; the highway was just changing names!

It shows promise, and has already been useful... but it's not fully cooked. Aside from little glitches, it really really really needs to let me select my route!
 
That must be one of the most generous reviews I've ever read.
"it sucked, it did this wrong, it got that completely wrong, it didn't do that thing - overall, it shows promise"
:-D

I think it depends on individual perspective.

Some people get frustrated with the trip planner that is still under works.

Some people take it easy and understand that it takes time to work out the issues.

My take is if they are inventing at lightning speed on other issues why are they finding it difficult to get a better trip planner and Nav ?
 
I guess the "End to Range Anxiety" has to work with what is. This is how it gets me the 511 miles to Rapid City from Bozeman, MT.
Either Choice still offers me some anxiety, or requires me to travel a bit out of my way. :) Not very practical, yet. Hope Wyoming gets it's SuperCharging act together sometime soon.
View attachment 78093

Honestly, at the same instant it pops up with that route it should cue up Take the Long Way Home from the Slacker Radio...
 
A 500 mile range pack would really end my range anxiety. I could do all the round trips I need to do without spending time charging, with range to spare and I could get to all my most needed Superchargers. Rapid City from Billings, SLC , Denver , Glacier National Park, Yellowstone National Park, Grand Teton National Park, The Beartooth Highway. All these right now require planning that 6.2 Still can't fix and a 500 mile pack will, someday.
 
Okay so here's one big problem with the algorithm. It calculates your route first, and THEN tries to add Superchargers.

As a result, the same algorithm that often has you do a round trip of 50 miles in the opposite direction of where you are going, also completely ignores Superchargers that are well within your range. And on a better route to your destination.

I've attached a screen shot from my recent trip. From Albany I can trivially drive to Utica, and from there drive directly home nonstop. Or I can go to Syracuse and up the Interstate. Instead, I get a route through the Cornwall Supercharger... which I can't even reach. Nutty result.

IMG_4042.JPG


Another nutty result is that I'll tell it to drive to a specific Supercharger that I can reach on one charge, and it will still try to backtrack me to a Supercharger in the opposite direction. Now sure, the estimated battery level at my destination was 20%, but that was an ample safety margin - and it was completely accurate by the way.

Another oddity is that my iPhone Google Maps application gives me the sensible routes that I actually want to drive as the default, yet the Google-based car navigation gives me stupid routes that are much longer and slower. I'm guessing only the address lookup is Google Maps, and the route selection is the built-in navigation system (Navigon or something).

So what we really need is a two-stage process. Either let the user drop some waypoint pins (best), or provide alternate routes to select from. Actually both, a la Google Maps, would be best. Only then, once you've selected a route, would the Supercharger search process engage.

Of course an even better solution would be to present several Supercharger-based routes first, and then let the user pick one or start dropping pins.

- - - Updated - - -

A major bug with the current software is its tendency to lock up. If you park the car for any reason (pick up a drink, stop to Supercharge), when you arrive back the system is showing a nutty result and doing an infinitely long recalculation.

As a recent example, I returned to the car after stopping for five minutes for some iced tea, and the navigation system is saying I'll arrive with -1% charge, and it's recalculating forever. Cancel navigation and refresh the address, and the route reappears and it correctly calculates a 21% reserve.

Another bug is the tendency for voice commands to mysteriously disappear. On three occasions during my trip I got to a decision point, and there were no instructions. In two cases I realized something was wrong and peered at the screen just in time to make the right choice. In one case I had already gone through the intersection before I realized maybe I'd gone the wrong way, but it turned out I was supposed to go that way after all.

On that topic, how about a mute on/off button right at the top interface level? Generally I only turn on voice instructions if I'm going through a complex series of turns.

If they could fix the bugs and routing issues this feature would really be very useful... as it stands it only works sometimes.

The energy app Trip screen is great... but is useless if the navigation chooses a different route that is way longer. Which happened three times during my trip. Fortunately Google Maps on my iPhone gave me the correct directions, but I didn't have a trip energy monitoring screen until the built-in Navigation clued into which way I was driving.
 
Another oddity is that my iPhone Google Maps application gives me the sensible routes that I actually want to drive as the default, yet the Google-based car navigation gives me stupid routes that are much longer and slower. I'm guessing only the address lookup is Google Maps, and the route selection is the built-in navigation system (Navigon or something).
Yes, I can pretty much guarantee that they are not using Google Maps to do the routing. On one trip that I do quite frequently a new road is missing in Navigon's database but is in Google. I can literally stand at one end of this 500m section of the road and route to the other end. Google Maps clearly shows the road but then the route created makes me drive a couple of miles around to the other end.
So what we really need is a two-stage process. Either let the user drop some waypoint pins (best), or provide alternate routes to select from. Actually both, a la Google Maps, would be best. Only then, once you've selected a route, would the Supercharger search process engage.

Of course an even better solution would be to present several Supercharger-based routes first, and then let the user pick one or start dropping pins.
The current trip planning is completely bogus and in my mind makes range anxiety worse not better. I just posted in the bugs thread that on my last trip I would have gotten stranded had I stopped charging when it told me to stop. That plus the ridiculous de-tours and back-tracking segments it adds... that's all just not useful. Why we are doing field pre-beta testing with the main customer base who might have gotten confused by Elon's press conference into thinking that this is actually production ready software is beyond me. I cannot envision a software development organization where the current trip planner would have met "start of beta testing" release criteria. Let alone "ship to customers".
Yes, I know it says "BETA" in the release notes. But how many people who buy a Tesla and just want to drive a car understand that this means "yeah, sure, play with it, but heck, don't trust it and don't drive where it tells you to go unless you verified this with a real navigation website, first"...
 
Yeah I understand it's beta, but it's really at alpha quality level. I don't mind testing it; in fact it was kinda fun. On occasion the notification that you're ready to go actually worked. On occasion. That was cool though.

If the feature worked 95% of the time it would be awesome. The problem is it works 10% of the time.

If it would allow you to manually add any Supercharger you wanted, and then recalculate the routing, it would be quite usable. Unfortunately it chooses which charging stations you're allowed to add, and they are never the ones I actually want to drive to. So I end up entering the just next Supercharger I want, and then deleting the extra one in the wrong direction it inserts.
 
On that topic, how about a mute on/off button right at the top interface level? Generally I only turn on voice instructions if I'm going through a complex series of turns.

I'm not sure what you mean by "at the top interface level", but since the most recent software update for me, there is a speaker icon button at the top left corner of the navigation interface that brings up the navigation volume/mute dialog, so one doesn't have to search for it at the bottom of the nav interface:

ImageUploadedByTapatalk1429549423.305025.jpg


Since the volume/mute dialog self-dismisses after 10 sec or so, it's a pretty solid UI improvement IMO. All you have to do is hit the speaker button and then the larger mute button to toggle it. Since that function only applies to navigation, it makes sense to place it on the navigation UI, and it was assigned the prime location there ...
 
Last edited:
Yeah I understand it's beta, but it's really at alpha quality level.

I see all of this as one small aspect of the much bigger problem. And that is that Tesla is just not spending nearly enough money on software development.

I just don't understand why they aren't. Tesla is so far out in front in so much of what they are doing, and in so many ways the software is limiting the possibilities, or turning what should be a spectacular experience for customers into merely a great or even just a good one. Tesla should strive to make every experience spectacular for every customer.

With mechanical components, that may not always be possible. There are always going to be issues with contactors and drive units and that stuff is somewhat unavoidable. To improve machining, etc. I'm sure is incredibly expensive. But you can improve the software by hiring one more good software engineer. So why not hire a hundred more? Or better yet, a thousand more!

When Tesla develops a great new battery technology it will only be useful, in all likelihood, for the new cars yet to be built. (If it is useful for cars already built, it will carry a large marginal cost.) When Tesla improves the software, on the other hand, much or all of the improvement can be rolled out to all cars, thanks to the innovative system Tesla developed for over the air software updates. So future owners as well as current owners benefit from software improvements. And once the software has been written, it's basically "free" to distribute it. I mean I know it's not really free, but the marginal cost of having to create another unit is essentially $0, unlike, say, an improved brake system or a better battery.

I honestly don't understand why Tesla isn't spending a boatload more money on software development.
 
I see all of this as one small aspect of the much bigger problem. And that is that Tesla is just not spending nearly enough money on software development.

I just don't understand why they aren't. Tesla is so far out in front in so much of what they are doing, and in so many ways the software is limiting the possibilities, or turning what should be a spectacular experience for customers into merely a great or even just a good one. Tesla should strive to make every experience spectacular for every customer.

With mechanical components, that may not always be possible. There are always going to be issues with contactors and drive units and that stuff is somewhat unavoidable. To improve machining, etc. I'm sure is incredibly expensive. But you can improve the software by hiring one more good software engineer. So why not hire a hundred more? Or better yet, a thousand more!

When Tesla develops a great new battery technology it will only be useful, in all likelihood, for the new cars yet to be built. (If it is useful for cars already built, it will carry a large marginal cost.) When Tesla improves the software, on the other hand, much or all of the improvement can be rolled out to all cars, thanks to the innovative system Tesla developed for over the air software updates. So future owners as well as current owners benefit from software improvements. And once the software has been written, it's basically "free" to distribute it. I mean I know it's not really free, but the marginal cost of having to create another unit is essentially $0, unlike, say, an improved brake system or a better battery.

I honestly don't understand why Tesla isn't spending a boatload more money on software development.
Maybe they're having trouble finding "qualified and compatible" people to hire. Send them a resume. ;)
 
Maybe they're having trouble finding "qualified and compatible" people to hire. Send them a resume. ;)

Heh. Not me. I can program my DVR, and I know my way around my Tesla menus, but I'm no programmer. I just know there are plenty of them out there. I respect what they do, and I fully expect that Tesla has hired some of the best and the brightest, and is just asking too much of them. So hire more!

I look at things like how long it took to have the very basics like shuffle, etc. added to the sound system software, or how long it took to add valet mode and those things just scream, "hire more people!"