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

Tesla Priorities: Refine Autopilot or Fix Everything Else?

This site may earn commission on affiliate links.
Mr. Musk needs to work for about five years in retail, dealing with the general public. .
+1
But since this won't happen there needs to be someone at a very high level who gets the importance of managing customer relationships. What boggles my mind is not knowing or seeing the face of the person who is ultimately handling the issues discussed above. Surely there is someone overseeing the day to day challenges while EM travels and focuses on hyperloop, space X, SCTY and going to Mars for heavens sake.
 
3 years ago, it was understandable that Tesla's software had limitations - the Model S was brand new - and they needed to invest resources to get the software more reliable and implement key features (like sleep mode). But after 3 years, that excuse really doesn't work any more - and if they want to have a lower priced Model 3, attracting a larger customer base - and will now be competing with an established competitor in the Bolt.

That's pretty much my problem with it. The user-facing side of the software does not represent 3+ years of development effort, in fact it shows stagnation. They want to compete with luxury cars (and quite still on the Model 3 end) this stuff needs severe attention.
 
I own a BMW i3 and I'm thinking of replacing my "long distance" Mercedes with a Tesla. It's threads like this that make me think I should hold off a few more months. Tesla not longer gets a free ride for being "new". The basics need to work entirely reliably. No beta, no maybe, no "oh it's just the xyz app, not critical". The basic stuff should all work. Audio, nav, heat, cool, phone, etc. It's scary to think that the same software development group that is responsible for auto lane change is responsible for the bugs in more mundane s/w.

I know all cars have s/w bugs. My 2014 i3 certainly has a few, but it's a $45K car, before incentives. I'm ready to buy a Model S, but for $100K, I expect working s/w, not betas, reliability, not bugs.

Is there any indication that Tesla has read this thread?
 
Last edited:
I own a BMW i3 and I'm thinking of replacing my "long distance" Mercedes with a Tesla. It's threads like this that make me think I should hold off a few more months. Tesla not longer gets a free ride for being "new". The basics need to work entirely reliably. No beta, no maybe, no "oh it's just the xyz app, not critical". The basic stuff should all work. Audio, nav, heat, cool, phone, etc. It's scary to think that the same software development group that is responsible for auto lane change is responsible for the bugs in more mundane s/w.

I know all cars have s/w bugs. My 2014 i3 certainly has a few, but it's a $45K car, before incentives. I'm ready to buy a Model S, but for $100K, I expect working s/w, not betas, reliability, not bugs.

Is there any indication that Tesla has read this thread?

BMW doesn't understand software, they have their own problems. But I do now appreciate that they did spend the time to get *simple* things right. Like how the wireless keys work, and day-to-day procedures. Like the heated steering wheel being buried in the menus would not fly at BMW. Totally trivial feature and implementation. I'm constantly shuffling settings in the car which should not be necessary.

If they read this thread I think they would spend the couple engineering days needed to fix it, no?
 
Tesla has a lot of work to do on previously released features. They need to focus on that for a while and stop introducing half baked new features to get media attention.
Agree, stop being like Microsoft and just adding new half baked buggy features, please fix the mapping application that has very significant problems and poor functionality and fix the media app. We've been waiting for a browser (even the current one) or Homelink in Australia for a year now, nearly 80km/hr sign is read as 30km/hr. This would be a brief amount of programmer time to address these issues, I understand that new development has been a priority but its time to fix the existing software as well
 
It's threads like this that make me think I should hold off a few more months.

But the coolest thing about Tesla, and not your precious Mercedes or any other manufacturer, is that Tesla will fix these bugs and update your car while you're sleeping one night and maybe provide even MORE features. This, to me, sets Tesla way above and beyond all manufacturers. Yes, there are glitches. But, they have the means to fix them without inconveniencing you to have to go into dealership.

The map comparison is the best example. Every year someone with GM calls me to "offer" me the upgraded maps on a CD for $150 (that need to be mailed to me). EVERY year, I ask them if the charging stations are a layer shown within the maps. EVERY year, they say no. EVERY year, I say thank you but no thank you. Tesla, on the other hand, uses Google Maps. When Google updates their maps, MS updates its maps wirelessly! Brilliant! :)
 
But the coolest thing about Tesla, and not your precious Mercedes or any other manufacturer, is that Tesla will fix these bugs and update your car while you're sleeping one night and maybe provide even MORE features. This, to me, sets Tesla way above and beyond all manufacturers. Yes, there are glitches. But, they have the means to fix them without inconveniencing you to have to go into dealership.

The map comparison is the best example. Every year someone with GM calls me to "offer" me the upgraded maps on a CD for $150 (that need to be mailed to me). EVERY year, I ask them if the charging stations are a layer shown within the maps. EVERY year, they say no. EVERY year, I say thank you but no thank you. Tesla, on the other hand, uses Google Maps. When Google updates their maps, MS updates its maps wirelessly! Brilliant! :)

Your point is only valid if they actually fixed the issues owners have been complaining about for years. The fact that they can fix it wirelessly doesn't matter, since I can just as easily go into the dealer with my vehicle and get a software update.
 
The map comparison is the best example. Every year someone with GM calls me to "offer" me the upgraded maps on a CD for $150 (that need to be mailed to me). EVERY year, I ask them if the charging stations are a layer shown within the maps. EVERY year, they say no. EVERY year, I say thank you but no thank you. Tesla, on the other hand, uses Google Maps. When Google updates their maps, MS updates its maps wirelessly! Brilliant! :)

Unfortunately, the hybrid navigation system (combination of Navigon maps & navigation software with Tesla's user interface and Google maps) is an example of what isn't working very well.

Yes, Tesla does distribute software updates periodically - and that is an advantage over other vehicles (though it's likely the other manufacturers will also start doing this too). And, we just received one of the promised "free" "annual" map updates for the navigation software.

However...

Since the Model S was introduced in 2012 (soon to be 4 years ago). The navigation software continues to lack many features present in other navigation systems. While the trip planner is a nice addition to help with trip charging - the software still lacks many features I had in my previous car (built around 10 years ago). We still can't do waypoints or route customization. The real-time traffic system doesn't provide any information on upcoming traffic issues. And the software still has bugs requiring periodic reboots due to screen freezes or loss of synchronization between the dashboard & console map displays.

While the Google maps are updated almost immediately after road changes are made - the navigation maps are only updated infrequently. Since getting my P85 at the beginning of 2013, I've only seen 2 map updates. And the latest map update appears to already be at least a year out of date - roads in our area built a year ago don't show up on the latest maps. And the inconsistency of the navigation and Google maps results in routes that are clearly not right - when there have been road changes recently. My street was built at least 7 years ago - and the navigation software thinks it takes 5 minutes to drive the first few feet on our block.

Because the AutoPilot software affects driver safety - they must continue to work on AP improvements as a top priority.

But Tesla surely can do more than one thing at a time - and should be able to also allocate enough resources to address the growing backlog of bug fixes, user interface issues, and missing functionality. That's what every other software development group does - invest resources in the newest features, while also continuing to support and maintain the previously released software.

And - because Tesla can quickly distribute updates, once they make those improvements - they can push them out to all of us...

Perhaps we'll start seeing more progress with software improvements with 7.2 and 8.0...
 
As the guy who started this thread, I feel good that I am not just some solitary whiner, and that people like Boatguy see that Tesla can't just get away with not fixing things under the excuse that the company is "new" -- because others are catching up or surpassing Tesla and the novelty and purportedly green aspects of an electric motor isn't enough to carry the day for many people. There will be consequences, and the company will feel it in the market, as Boatguy indicates. Volvo's new interface is really nice, and is also upgradable OTA for enhancements -- and their new version of autopilot is very intriguing. Most everyone has better voice controls than we do at this point, and even Chevy has Apple's car interface.
Yes, ChriSharek it is cool that they can fix these things instantly with OTA updates, but as Zambono says, that is only good if they do it -- which we all assumed they would when we bought in, but they haven't fixed a lot of this in three years. Hopefully they are reading this thread, but you can help by sending the link with a one or two sentence intro to the Tesla service email, where they supposedly take our suggestions.
 
While I've greatly enjoyed my 2012 P85 and working with the Tesla phone support and local service center has been fantastic, Tesla's underperformance on software quality and functionality may be a factor in our decision to purchase another Tesla.

And Tesla's inability to provide expected functionality in the core apps and the presence of obvious bugs & design flaws (time/temperature location, periodic reboots, ...) does raise concern about being able to trust Tesla to deliver more complex software like AutoPilot - where even small software flaws could lead to accidents.

As someone who has been responsible for developing major software systems, I know it is possible to consistently meet aggressive schedules, while also delivering the functionality users expect - and fixing major bugs before the software is released. But this doesn't happen by itself - it takes management commitment to build a culture, recruit a team and craft a development strategy that can produce high quality - high functionality releases - on time.

And, the longer Tesla waits to address these software development issues - the harder it is going to be to get this under better control - and the greater the risk that another manufacturer will start building competitive long range EVs - and when that happens, the onboard software could be a significant drag on new sales.

Instead of announcing the next shiny new major software feature - what would be even better was a strong commitment from Musk down through the organization - to do better with their software - address the growing backlog of user concerns, change their software process to detect more bugs before software is released, and to fix the frustratingly random software distribution process & provide owners more information & control over the software running in their cars.

Hopefully Tesla is addressing these issues - and we'll see noticeable progress in the next releases...
 
While I've greatly enjoyed my 2012 P85 and working with the Tesla phone support and local service center has been fantastic, Tesla's underperformance on software quality and functionality may be a factor in our decision to purchase another Tesla.

And Tesla's inability to provide expected functionality in the core apps and the presence of obvious bugs & design flaws (time/temperature location, periodic reboots, ...) does raise concern about being able to trust Tesla to deliver more complex software like AutoPilot - where even small software flaws could lead to accidents.

As someone who has been responsible for developing major software systems, I know it is possible to consistently meet aggressive schedules, while also delivering the functionality users expect - and fixing major bugs before the software is released. But this doesn't happen by itself - it takes management commitment to build a culture, recruit a team and craft a development strategy that can produce high quality - high functionality releases - on time.

And, the longer Tesla waits to address these software development issues - the harder it is going to be to get this under better control - and the greater the risk that another manufacturer will start building competitive long range EVs - and when that happens, the onboard software could be a significant drag on new sales.

Instead of announcing the next shiny new major software feature - what would be even better was a strong commitment from Musk down through the organization - to do better with their software - address the growing backlog of user concerns, change their software process to detect more bugs before software is released, and to fix the frustratingly random software distribution process & provide owners more information & control over the software running in their cars.

Hopefully Tesla is addressing these issues - and we'll see noticeable progress in the next releases...

I wonder why this is not important to them. Musk comes from software. He must know how important this is. Would be interesting to know what is happening behind the scenes. Do they just not have the resources to put towards the software issues?
 
As one who traded up soon after I drove a loaner for a couple of days that had TACC. AP was delivered shortly after receiving our 70D. I certainly approve of their software approach and urge you to fear not. AP beta is fantastic and will only get better. A test drive will show you what we are talking about. Next upgrade I expect to see is v2 of sensor package (probably software they are developing for Model III). I expect to see v2 sensors and compute platform on next generation S/X platforms about a year before III deliveries. Similar to V1 sensors on the S came out about a year before X roll out. Just as V1 sensors were not backward compatible to the classic S, V2 will not be compatible with current S/X platforms. Also expect to see autonomous AP about 2-3 years further down the road, once again new sensor suite for new platforms.
 
Right. It isn't about "either - or", nor if Tesla's ability to deliver OTA updates is a possibility, nor if Tesla is intently focused on Autopilot and continues to deliver updates to that functionality, nor if there will be a V2 of Autopilot hardware on future vehicles that NO current owners have. What this is about is, Tesla and Elon's choices. Elon is in charge, by his decision, indecision, or his delegation. Such is life as the leader.

What just isn't being done is addressing acknowledged software bugs, beefing up Nav to be competitive, and fixing the horrible Media Player problems and deficiencies many of us encounter on a daily basis. Others have recently articulated the challenges Tesla WILL have if they are not already on top of this and a present is coming to us in an imminent OTA early Christmas present.

I just returned from a several-day road trip -- my first after having taken delivery, and while I L-O-V-E the mechanics of my MS, the SpC Network and how that operated for me -- even while the trip took another 90 minutes or so on each end because of time-to-charge vs. just fill-up with Dino-fuel in my previous hybrids, I truly missed having basics like Nav waypoints so I could put in my final multi-day destination (and SpC locations I planned to stop at on the way) like I have with my former luxury vehicles for years; having to remember to not leave my USB music on or muted when I gave my keys to the valet to avoid phantom playing and additional range loss while my MS was out of my control (being charged -- yeah, a GREAT perk at the right destinations!), and watching the evolving stupid and inappropriate (no longer funny) Gracenotes album covers for hours pop up in front of me on the IC and 17" when I know the album art I have tagged to each of my 16K tracks is perfect and displays properly on every one of my iTunes devices. The little things take away from my otherwise superb Tesla ownership experience, and after just 5 months of ownership, WILL be the reason I won't buy another Tesla when that day comes if things don't change to meet my expectations. I appreciate these things may not be significant for some of today's cash-rich enthusiasts that care about only performance or being part of a movement, but it is for those of us that have owned competive vehicles that have come to expect such refinement as an imperative. As has been well said, Tesla no longer gets a pass as a "start-up" on these things, and they can't focus perhaps only on Elon's future vision to the exclusivity of what the next layer of Tesla owners expect.
 
Right. It isn't about "either - or", nor if Tesla's ability to deliver OTA updates is a possibility, nor if Tesla is intently focused on Autopilot and continues to deliver updates to that functionality, nor if there will be a V2 of Autopilot hardware on future vehicles that NO current owners have. What this is about is, Tesla and Elon's choices. Elon is in charge, by his decision, indecision, or his delegation. Such is life as the leader.

What just isn't being done is addressing acknowledged software bugs, beefing up Nav to be competitive, and fixing the horrible Media Player problems and deficiencies many of us encounter on a daily basis. Others have recently articulated the challenges Tesla WILL have if they are not already on top of this and a present is coming to us in an imminent OTA early Christmas present.

I just returned from a several-day road trip -- my first after having taken delivery, and while I L-O-V-E the mechanics of my MS, the SpC Network and how that operated for me -- even while the trip took another 90 minutes or so on each end because of time-to-charge vs. just fill-up with Dino-fuel in my previous hybrids, I truly missed having basics like Nav waypoints so I could put in my final multi-day destination (and SpC locations I planned to stop at on the way) like I have with my former luxury vehicles for years; having to remember to not leave my USB music on or muted when I gave my keys to the valet to avoid phantom playing and additional range loss while my MS was out of my control (being charged -- yeah, a GREAT perk at the right destinations!), and watching the evolving stupid and inappropriate (no longer funny) Gracenotes album covers for hours pop up in front of me on the IC and 17" when I know the album art I have tagged to each of my 16K tracks is perfect and displays properly on every one of my iTunes devices. The little things take away from my otherwise superb Tesla ownership experience, and after just 5 months of ownership, WILL be the reason I won't buy another Tesla when that day comes if things don't change to meet my expectations. I appreciate these things may not be significant for some of today's cash-rich enthusiasts that care about only performance or being part of a movement, but it is for those of us that have owned competive vehicles that have come to expect such refinement as an imperative. As has been well said, Tesla no longer gets a pass as a "start-up" on these things, and they can't focus perhaps only on Elon's future vision to the exclusivity of what the next layer of Tesla owners expect.

Good post. Since this was your first longish trip, how much if any of that 90 minutes could you attribute to charging more than necessary due to range anxiety and therefore have shaved off?
 
Good post. Since this was your first longish trip, how much if any of that 90 minutes could you attribute to charging more than necessary due to range anxiety and therefore have shaved off?
I'll put it this way. I used EVTripPlanner to get a feel for options before I left. That helped a lot with contingency planning and easing my initial concern. Without it, I wouldn't have been nearly as comfortable. I just don't want to sit in my MS to do that sort of planning I'd rather do at my desk. Plenty of charging options on the 1/3 of the trip nearest to my home, so I wasn't too worried, I'm just a planner, so like to have a worked-out Plan B in my hip pocket and a Plan C at least half-baked. The big CA storm was approaching and weather people were predicting the worst here, so I decided to travel half-way on Saturday afternoon to hopefully stay ahead of the storm (that didn't really come here, but hit me unexpectedly the next day.)

It didn't cost me any time or really much money to trip-charge to 100% before I left home... First leg of the trip went fine without a lot of weather, but had several challenges with traffic where Nav rerouted and I made it to my new overnight stay next to a SpC easily with more than 25% charge left. Plugged into the SpC, checked-in, unpacked, went to dinner, and came back to have charging at 90% just about done. Moved MS over to the hotel and parked.

Next morning, had unpredicted rain -- sometimes causing a slowdown on the freeway to see -- and 30-40mph crosswinds going, so I was more uneasy. Made it 2 hours later to the next SpC with that bad weather the whole time I was traveling 65-70 for the most part, took a nature break, bought another cup of coffee to allow my MS to finish charging to a full 90% just in case. It wasn't absolutely needed, but that was sort of a strategic decision to charge and know I'd have enough to get to my destination and return back to that SpC if I didn't get a charge while I was where I was headed to. The valet ended-up charging my MS while I was parked (they couldn't guarantee it, but they had IIRC 4 80A HPWC which is great with my dual-charger). I got the indication it was charging on both my Apple Watch and iPhone the day after I left my MS with the valet, so was pretty stoked to see it all working as designed.

Unfortunately, after several hours after my MS finished charging, they moved my MS out of LTE range in the depths of the garage somewhere, so I couldn't trip charge to 100% when I woke up this morning as I was going to do via my iPhone. Regardless, was able to pretty much skip the intermediate SpC on the at home (only plugged in while I was doing a little nature break and getting a cup of coffee since there were 8 open stalls, so why not opportunity charge for 20 mins?). At the 2nd stop, it took a little over an hour to put 90% back into my MS for the longer leg home while I had breakfast at Denny's and checked email. Made it home at 60-75mph and loads of stop-and-go traffic with about 25% charge left when I pulled into my garage... Plenty to then go pick up my dog from camp, have enough reserve should an emergency come up with family, and plug back in for charging at midnight.

Net is, trip worked-out fine. I bet maybe 30 minutes of each side of ~90 minutes could have been shaved off, but it's hard to estimate what weather will and won't do, and I'm still worried when I'm out in the middle of the desert vs going from metropolitan area to another where there is more infrastructure... Same deal e.g. with people I suspect in the Midwest and center of Canada... Cross winds seem to be the killer, rain didn't help. Amazing to watch the displays and impact on product range while all that wa going on and still needing to climb significant elevation. I had no anxiety with 25% remaining on the trip graph, but slowed it down a bit when it projected less. Great weather, I won't need to be as big of a buffer for sure. EVTripPlanner was very accurate (albeit using an 85D as the baseline), including accounting for very significant difference in altitude gain (3K'+) on the outbound leg vs return, and therefore difference in range per charge... Larger concern of course is what if one of the SpC I'm dependent upon 2+ hours away is offline when I get there, so I'm glad the later versions of firmware have some indication in that regard. Not completely trusting Tesla's attention to detail though, I admit to trying to double-check alternate sites before I left each place just to add to my confidence level. Yes, I'm paranoid about being stranded by myself in the middle of nowhere with speeding passer-bys that wouldn't stop if I did need help.

...and back to an adjoining tread topic, I'm also frankly still happy to have my supposed "16 miles"... It costs a lot, and I'll never really know of its all there, but it is one more thing to ease range anxiety for those of us that sometimes travel the wide open spaces in the wild and wooly West -- with only rattle snakes, semi's, and dead cow skulls between SpCs. ;)
 
Last edited:
Recently, I was driving someone who had never been in a Tesla before. Of course the large console is very impressive - and the software looks great (as long as you don't look too closely and try to really use it).

But when the display locked up - and I had to reboot the software while the car was in motion, well, it wasn't very impressive...
 
Further to the question of whether Tesla would be better off putting all their software development chips on red (autopilot) or using some of that capacity trying to round out and perfect the functionality of the UI and UX, it looks like most other auto manufacturers are catching up on the self-driving game faster than Tesla anticipated: A $20,000 Self-Driving Vehicle Hits the Road.
The UI and UX is still a great potential differentiator for Tesla -- it would be a pity to blow this, especially with the Model 3 no more than two years away (we hope).
 
Further to the question of whether Tesla would be better off putting all their software development chips on red (autopilot) or using some of that capacity trying to round out and perfect the functionality of the UI and UX, it looks like most other auto manufacturers are catching up on the self-driving game faster than Tesla anticipated: A $20,000 Self-Driving Vehicle Hits the Road.
The UI and UX is still a great potential differentiator for Tesla -- it would be a pity to blow this, especially with the Model 3 no more than two years away (we hope).