Tesla Priorities: Refine Autopilot or Fix Everything Else?

[RichardL]

AR +100
Reading this thread (and the other similar ones) I get the feeling that many (myself included) heard what was implied by Elon's tweet about features to go in 7.1 and are deeply disappointed by the PR version we actually got. It feels like something fundamental has changed in our attitude.

 

[Cyberax]

Autopilot, automatic summoning, and other nifty features are lost on me. This does not mean that I think they should go into the scrap pile. I am sure that there are valid safety and convenience reasons why these "luxuries" (for lack of a better term) are being introduced. But I question why they are so important now. Moreover, I really do not know if any or all of these features are standard on all makes and models or only on a selected few.

For better or worse, a lot of people these days buy cars based on the media/infotaiment features.

And it actually makes sense, if you think about it. Pretty much all cars on the public roads travel at the same speed - your 40 minute daily commute will take 40 minutes on any car. It doesn't matter if you have a Hellcat or a lowly Honda Civic. So as a result most regular buyers don't really care about stuff like V6 vs. V8 and turbopumps vs. naturally aspirated engines.

Ditto for regular trim options - even lowly cars these days have electronically adjustable seats, autodimming mirrors and multi-zonal climate control. There's little left to differentiate.

Infotaiment and assistance systems are another thing - it makes sense to buy a car with a good infotaiment system since you're likely going to be using it constantly.

 

[calisnow]

I argue that if the purpose of Tesla enhancements is to sell more cars, rather than to pump the stock price or create a revolution (debatable, I know), that Tesla should be focussing on optimizing the overall user experience with the car, and just hang in the hunt on auto pilot, since the best car companies are all doing it anyway, and this will not be a differentiator of Tesla from other cars a year from now

I sympathize with you, and understand it must be frustrating to have a UI in your car that you do not find optimal. However I would make two observations: 1 - I imagine Tesla knows better than you what will sell more Teslas. 2 - You have absolutely no idea what the autonomous car market will look like in the coming years - you are an outsider writing on an internet forum (as am I). The history of AI going back to the 1940's is to assume that whatever challenges remain are simple and will be fixed shortly. Just because everyone is discussing self driving and researching it does *not* mean that everyone will have it shortly. For all you know the challenges in expanding autopilot's capabilities are deeper than they appear and require more resources to develop than you think they do. And 3 - Self driving is in my opinion, as a person who has two Teslas currently on order - absolutely the most important priority I want Elon to focus on. Sorry man but if forced to choose between a better internet radio interface and a car which will more reliably make the right decisions when driving itself down the road - I'll take the improved self driving.

 

[brooklynrab]

I sympathize with you, and understand it must be frustrating to have a UI in your car that you do not find optimal. However I would make two observations: 1 - I imagine Tesla knows better than you what will sell more Teslas. 2 - You have absolutely no idea what the autonomous car market will look like in the coming years - you are an outsider writing on an internet forum (as am I). The history of AI going back to the 1940's is to assume that whatever challenges remain are simple and will be fixed shortly. Just because everyone is discussing self driving and researching it does *not* mean that everyone will have it shortly. For all you know the challenges in expanding autopilot's capabilities are deeper than they appear and require more resources to develop than you think they do. And 3 - Self driving is in my opinion, as a person who has two Teslas currently on order - absolutely the most important priority I want Elon to focus on. Sorry man but if forced to choose between a better internet radio interface and a car which will more reliably make the right decisions when driving itself down the road - I'll take the improved self driving.

If you want the most reliable and polished self driving package 3 years from now, I'll bet you that you will be looking at a BMW or Mercedes, not a tesla. They have the resources to bury tesla on this, they are focusing on it now, and they (unlike tesla) have the experience in perfecting a thoroughly refined product. My point is that few people are going to buy one car over another because of perceived minor differences between their auto pilot suites. If tesla is perceived as being among the class leaders, that will be plenty good. The differentiator for this car, in addition to the electric motor, is the similarity to and utility of your smart phone. Which means making the user interface as perfect and simple to use as possible, starting with nav and voice command.

 

[Cyberax]

If you want the most reliable and polished self driving package 3 years from now, I'll bet you that you will be looking at a BMW or Mercedes, not a tesla.

Quite a few C?O types I know actually looked at Tesla.

My point is that few people are going to buy one car over another because of perceived minor differences between their auto pilot suites. If tesla is perceived as being among the class leaders, that will be plenty good.

Actually, I for one bought a fully-loaded-except-P Tesla just because of these minor AP features and promises of future central console upgrades.

The differentiator for this car, in addition to the electric motor, is the similarity to and utility of your smart phone. Which means making the user interface as perfect and simple to use as possible, starting with nav and voice command.

Except right now Tesla has worse voice commands and navigation than my 3 year old Chevy Volt. If you insist on comparing Tesla with cellphones then it's a phone from a bargain bin at Walmart checkout line, not a sleek new iPhone 7.

 

[ohmman]

The differentiator for this car, in addition to the electric motor, is the similarity to and utility of your smart phone.

I would argue that the differentiator is simply the electric drivetrain, the knowhow that has gone into creating that drivetrain, and maybe a head start on a serious charging infrastructure. The utility of your smart phone and user interface aren't differentiators if autopilot isn't. You can't cherry pick one without selecting the other. The thing Tesla has on everyone is a dedication to an electric drivetrain. Other manufacturers will continue to be distracted by their core businesses - until they're not. I don't know when that'll be, or if another pure electric entry (Apple?) will come to market. But that, and nearly only that, is the thing that separates Tesla from the alternatives.

When we ordered our MX, it wasn't because I think that the MX body, UI, doors, etc. have anything over the shell of the competition. In many ways, I think it lags. The only reason I ordered was because of the drivetrain. In a lot of ways, I did that in spite of the drawbacks. I want an electric car that I can take on road trips. I want my garage to have no ICE vehicles in it, and I want to be done with gas stations and oil changes. It's nearly 2020. I expected flying cars. This is the closest I can currently get.

I like the idea of giving Tesla feedback on what people would like to see. I don't like the idea that Tesla should be focused on legacy vehicles like my classic MS, though. They should focus on properly supporting my vehicle (and definitely fixing Navigation), but new features are unnecessary. I wasn't promised more voice features, and I'll still be satisfied if I don't get them. The car was more than good enough for me when I took delivery of it, and it continues to exceed expectations.

 

[byan1232]

I agree with you. As I suggested earlier, I would be OK with a slow but sure approach to knocking some of this off with perhaps an internal team who owns the non-Autopilot/Performance capabilities -- while a majority perhaps continues work on the jazzy new stuff that is needed to attract the Press and some new buyers at the same time. There is little (note I didn't say "nothing") to demonstrate Tesla Mgmt is focused that way.

As a software guy many years ago, I was hired out of college into the "maintenance team" who's job it was to fix code that was broken and do minor enhancements as the need arose... (I was the guy that got the call at 2AM when e.g. Payroll blew up, or there was some new terms in a negotiated union contract to integrate with Time & Attendance). The more senior "development" teams built new capabilities and new releases that once accepted by the user community, was turned back over to "maintenance". Both had to work in a coordinated manner, but it allowed future investments to move forward, at the same time what was running the business kept working 7x24x365 with any minor fixes and a few improvements were slipped in along the way. It does not appear Tesla operates like this, and I know a lot is different in the 35+ years since my programming days, but don't you agree the old concept sounds like something that is a lot more customer-centric than perhaps is being delivered today? ​

I'm a relatively new MS Owner, but i must say I find it a little disillusioning to find some of the problems and usability issues I've encountered are the same as have existed from Day 1, with now 100K MS on the road -- especially when resolution will be via software only and can be distributed OTA. I do love my MS. I also appreciate where Elon is likely headed into the future, but he and Tesla must have more attention to these littler details as Tesla grows and matures. The unresolved problems and little nits add up with some of us, and at least for me, my love for my MS becomes ever-so-slightly less as time goes on and nothing seems to be done to resolve issues I have every time I drive my MS. I'm sure if Tesla does not gain focus on doing more "maintenance" as well as future "development" at the same time , it will catch up with them in the longer run. "The masses" will expect more, yet pay less, than some (or is it "many") earlier enthusiasts have and will.

Hey what little problems are you having that's annoying? I'm awaiting delivery of my model s and would love to know more about the car.

 

[calisnow]

If you want the most reliable and polished self driving package 3 years from now, I'll bet you that you will be looking at a BMW or Mercedes, not a tesla. They have the resources to bury tesla on this, they are focusing on it now, and they (unlike tesla) have the experience in perfecting a thoroughly refined product. My point is that few people are going to buy one car over another because of perceived minor differences between their auto pilot suites. If tesla is perceived as being among the class leaders, that will be plenty good. The differentiator for this car, in addition to the electric motor, is the similarity to and utility of your smart phone. Which means making the user interface as perfect and simple to use as possible, starting with nav and voice command.

The differentiator for this car *to me* is not having the interface be as slick as an iphone - the interface is already great (I've spent 2 weeks renting two different Teslas before I ordered - one with no AP and one with). The differentiator is that it is the best damn driving car I have ever known and it doesn't use gas. And I'm not coming to Tesla from boring cars - I've been a sports car nut my entire life and have owned a string of German and Italian iron (still have the Italian iron and the Tesla will be replacing an AMG Benz). The Tesla is more fun to drive than all of them (with the exception of the Italian, but that isn't relevant competition).

As for where you think the rest of the industry will be with autopilot in 3 years - we shall see. One data point right now is in the February 2016 edition of Car & Driver (or Road & Track, I forget) - did you read the article? The BMW and Benz current in-production driver assistance systems both have error rates over 100% higher than Tesla's system, despite having more sensors.

Will everyone improve rapidly? Of course. But it is possible that Tesla's massive on-road beta testing fleet is giving them a first mover advantage that will get them to market with TRUE autonomous go-to-sleep driving several years earlier than competitors can. And if so that will be ANOTHER differentiator. I don't know about you, but if I can take a nap while my Tesla drives me 300 miles north from L.A. to Mammoth to go skiing I will upgrade in a heart beat.

Also - these two Teslas are the first cars I've leased in forever. I never lease - stupid financial decision. I keep cars til the wheels fall off - typically over 200,000 miles for daily drivers. But Teslas are improving so ridiculously fast that I don't want to get left behind.

You people can keep whining about the half-baked software features - the larger buying public doesn't care. It's a tempest in a tea cup for a bunch of geeks like us who spend our time arguing on the internet. P

Three years from now both the software and hardware will be more fully baked. Their dev teams will be deeper as they grow. The range will be longer, the cars will have more autonomous capability.

Neither you nor I have any idea what the competitive landscape will look like, and this conversation will be a forgotten memory.

In the mean time Tesla's sales continue to show explosive growth while we sit here being keyboard jockeys. They're doing just fine.

And if Tesla has been dethroned 3 years from now in the autonomous driving/efficiency/handling/UI area - well then free markets will have done what they always have done - improved the product. You'll be able to get something even more magnificent to drive than the Model S you have now.

- - - Updated - - -

Autopilot programming and general UI programming require almost completely different skillsets. Internally they should have at least two teams, one doing the AP stuff and another doing the UI design. And a third one doing the low-level microcontroller programming.

Perhaps they do?

 

[ecarfan]

I would argue that the differentiator is simply the electric drivetrain, the knowhow that has gone into creating that drivetrain, and maybe a head start on a serious charging infrastructure...I want an electric car that I can take on road trips. I want my garage to have no ICE vehicles in it, and I want to be done with gas stations and oil changes. It's nearly 2020. I expected flying cars. This is the closest I can currently get.

My thoughts exactly. I am so pleased to own two Teslas and know that I will never visit a gas station again. I'm driving the future...today. Sure my S isn't perfect, and it has no AP, but it's a much improved car over when I bought it two years ago. That's astonishing! No other car manufacturer has ever done that.

 

[David_Cary]

Ok - I'm a relative newbie having started with 6.2 in 5/15 so I don't get the level of frustration you have - and I suspect the thread title has brought out a certain position. I have very little issues day to day and I am nowhere near frustrated. My only issue is with the media browser needing to hit FF sometimes to get a song to load from a playlist.

Nav - works pretty good for me. Better than other cars. Voice recognition for location entry - nearly perfect and better than other OEMs that I've had. Essentially as good as Siri for me. Time prediction and traffic routing - certainly better than iOS but not better than Google. No OEM can match Waze or Google. If anyone could, it would probably be Tesla allowing us to use those two. Waypoints? I couldn't care less. The supercharger thing is ridiculous.

Improvements for me have been great. Calendar function is slick. Auto homelink works great. TACC - even something out less than a year has improved markedly.

Ok - there is lots of room for improvement. But honestly so is every other software I've ever used. I use software every day that sucks - I mean truly terrible. I hate to say it but I have very few good things to say for the entire field of software development. Anyone in medicine will agree with me (I think???).

 

[davidc18]

We like our car and we are going to replace our second car (honda odyssey) with a second MS. With that said, I have to agree that outside of the drivetrain and TACC/AutoPilot the car could use some attention.

1. browser - seriously outdated
2. music player - as basic as it comes
3. nav system and maps - how about updating the browser so we can use google maps?
4. charging system - why can't you set a time to have a set charge level?

anyway, I now understand that Tesla will devote Zero resources to finishing these products, which is sad.

 

[511keV]

The Tesla Model S has been on the road for three years and I've had mine for 7 months but I still consider mine a Beta status vehicle. Tesla's 7.1 release was a non-event for me and still leaves the car in Beta status. I see half finished projects not only with Autopilot, stable NAV routing, and kludgy UI but the sparsely populated Apps tab of my center console. Where is the application API? While there might be a NAV improvement, the much maligned Apple IOS Map app is far ahead. I've come to accept better integration with smart phones via Apple CarPlay (or its Goggle equivalent) will probably never happen and I'll soon be consigned to be a habitual law breaker if proposed laws are adopted in Massachusetts for having to handle my iPhone to perform some function that should be controllable via voice commands, steering wheel controls, or center console. Also, Tesla consistently turns a deaf ear to features that people in non-California coastal climates (aka climates that regularly go below 0 C) such as scheduling charge end time and decoupling battery heating from range mode while connected to a 240V circuit desperately need to have a warm battery allowing the car to drive consistently with regen.

 

[byan1232]

decoupling battery heating from range mode while connected to a 240V circuit desperately need to have a warm battery allowing the car to drive consistently with regen.

Sorry if this is a stupid question but what do you mean by this? I live in Boston so when I receive my model S in March, I would like to know as much as possible.

 

[BertL]

Hey what little problems are you having that's annoying? I'm awaiting delivery of my model s and would love to know more about the car.

Specifics are listed in many other threads (do searches and you'll find the discussion). For me, there are two lists -- one is what's broken or significantly missing compared to my former Lexus, BMW and MBZ -- and another set of items on my improvement wish list I hope someday to receive as an unexpected gift. To a large degree I knew what my MS came with when I placed the order, so I try not to gripe as much about new features and improvements that I'd love to have some day. What I do complain about are things clearly broken, not operating consistently, or are what I consider basics that I had and do make use of on a regular basis, e.g.

• Safety issue with 7.x UI choice to move Speed, TACC, Autopilot and all "idiot lights"/indicators to top of the IC -- It's a significant issue for me since after I've moved the steering wheel to an even higher and much less comfortable position since I took delivery with original 6.2 firmware, I still can't easily see the most important things on the IC and possible warnings needed to safely drive my MS. (I'm 5'11" and others here have said they have the same problem. See representative photo below.)
• 90kWh Battery has inaccurate Rated Range -- Tesla-acknowledged defect -- see other threads on the long detail
• USB Media Player Support -- Several Tesla-acknowledged defects. Of particular note for me, is if I play USB music (which I do all the time) it will sometimes continue to play when the car is off - further draining my battery by 2-4 mi/day. It also makes listening to longer tracks like audiobooks nearly impossible since there is no way to stop and ensure MS will pick up playing where YOU stopped listening; Occassional "unable to play" message upon restarting MS; No/inconsistent use of personal or Gracenote Cover Art like other music sources -- instead Tesla presents a huge grey static (and very annoying) box on the IC and 17"; No playlist support (I wrongly assumed this would just be there when I purchased, since Tesla elected not to provide native iPod support like most other brands)
• Driver Windshield Wiper Chatter - Tesla-acknowledged defect (yes, I know this is hardware not software, but the problem has been there since MS was first produced.)
• Navigation needs Waypoints; Alternate route choices (e.g. time vs distance); Ability to exclude toll roads, etc in route selection. These were not there when I ordered my MS, but honestly are basic usability capabilities in nearly every other non-luxury vehicle with OEM NAV the past several years, and it existed in my previous Lexus and MBZ which my MS replaced.

Other people could chime in with other things that are broken and bug them (cold feet/heating, etc), but lets not go too far off topic. There are plenty of other threads with those debates, but hopefully byan1232, that gives you a feel for my top-of-mind concerns.

 

[msnow]

Ok - I'm a relative newbie having started with 6.2 in 5/15 so I don't get the level of frustration you have - and I suspect the thread title has brought out a certain position. I have very little issues day to day and I am nowhere near frustrated. My only issue is with the media browser needing to hit FF sometimes to get a song to load from a playlist.

Nav - works pretty good for me. Better than other cars. Voice recognition for location entry - nearly perfect and better than other OEMs that I've had. Essentially as good as Siri for me. Time prediction and traffic routing - certainly better than iOS but not better than Google. No OEM can match Waze or Google. If anyone could, it would probably be Tesla allowing us to use those two. Waypoints? I couldn't care less. The supercharger thing is ridiculous.

Improvements for me have been great. Calendar function is slick. Auto homelink works great. TACC - even something out less than a year has improved markedly.

Ok - there is lots of room for improvement. But honestly so is every other software I've ever used. I use software every day that sucks - I mean truly terrible. I hate to say it but I have very few good things to say for the entire field of software development. Anyone in medicine will agree with me (I think???).

I think the NAV/routing must be geographical thing because if you lived where I do NAV will misroute and get lost 6 out of 10 times. You wouldn't like it. Agree about the voice recognition being good but I think the OP's issue with it is they want more than just 4 or 5 commands.

 

[byan1232]

Navigation needs Waypoints; Alternate route choices (e.g. time vs distance); Ability to exclude toll roads, etc in route selection. These were not there when I ordered my MS, but honestly are basic usability capabilities in nearly every other non-luxury vehicle with OEM NAV the past several years, and it existed in my previous Lexus and MBZ which my MS replaced

Wow, I agree Tesla should focus on issues like these. I thought the NAV uses GoogleMaps so shouldn't it have the same capabilities as GoogleMaps?

 

[msnow]

[*]Driver Windshield Wiper Chatter - Tesla-acknowledged defect (yes, I know this is hardware not software, but the problem has been there since MS was first produced.)

During the rain storms we just had I paid special attention to see if my car had any wiper chatter and I can say it didn't. Unless there are some special circumstances that cause this to happen I think you should revisit this topic with your SvC because our cars were built around the same time so I assume they have identical hardware.

 

[msnow]

Wow, I agree Tesla should focus on issues like these. I thought the NAV uses GoogleMaps so shouldn't it have the same capabilities as GoogleMaps?

No, it uses Garmin/Navigon and hasn't been updated in over two years.

 

[ecarfan]

I recall the Navigon map data being updated sometime in 2015.

 

[stevezzzz]

Sorry if this is a stupid question but what do you mean by this? I live in Boston so when I receive my model S in March, I would like to know as much as possible.

If you leave your car in Range mode, you get the range benefits of more aggressive torque sleep and a limited draw for climate control while driving. But when you're plugged in to 240V shore power at home, overnight, Range mode turns off or severely limits battery heating during pre-conditioning of the interior (which many people do using the smartphone app, before unplugging in the morning), which kind of defeats the purpose of Range mode, because without a warm battery you will find that regen braking is severely limited or non-existent when the ambient temperature is low.