TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker or making a Paypal contribution here: paypal.me/SupportTMC

Why AP 2.0 Won't Be Here Soon, and It Won't Be What You Think It Is

Discussion in 'Model S' started by Todd Burch, Sep 12, 2016.

  1. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    #1 Todd Burch, Sep 12, 2016
    Last edited: Sep 12, 2016
    There is a lot of speculation that Autopilot 2.0 is imminent--perhaps coming within a quarter or two. There's also a lot of speculation that Autopilot 2.0 will be capable of level 4 autonomy.

    I'm going to put myself out there and say: Not a chance.

    I should "color" this opinion, as the financial analysts like to put it, by stating that:
    1. I am a huge Tesla fan
    2. I am a huge Elon Musk fan
    3. I am a fan of technology
    4. I am a happy early Model S adopter (reservation January 2010, first S delivered Dec. 2012).
    But this forum tends to be the king of manufacturing unrealistic expectations. I'll summarize this post right off the bat by saying the following: If you are putting off a Tesla purchase waiting for autonomous driving, you're wasting your time.

    You will probably not see AP 2.0 hardware for another few quarters at the very earliest. And that's being optimistic.

    You will not see AP 2.0 (what I am using to characterize level 4 autonomy) for about 5 years. And that's being VERY optimistic, even keeping in mind the blindingly fast pace Tesla is moving with this technology.

    You will not see level 3 autonomy from Tesla for at least 2 years. And that's being optimistic.

    Let me explain.

    Autopilot has been out for a year now. While improvements have been remarkable, think about the basic highway driving scenarios that aren't handled yet.

    1. Every time I crest a hill, my car dives for the left or right side of the road unless I'm following another car.
    2. Every time the lane markings fade, the car drifts and I have to take over.
    3. Every time I pass an entrance ramp with cars merging, I have to take control. AP does not handle sequencing itself with merging vehicles.
    4. Every time I'm merging myself, I have to take control. The car cannot sequence itself to merge onto the highway.
    5. The car does not automatically change lanes to maintain a target speed.
    6. The car does not avoid large road obstructions.
    7. The car does not move laterally to avoid parked cars on the shoulder, bicyclists, or pedestrians.
    8. The car is not always clear about which lane a car ahead is in. Sometimes TACC slows for a car in an adjacent lane.
    9. If a car cuts in front of you, the braking is later than most would consider to be comfortable.
    10. The car cannot stop as smoothly or gradually as a human would (though something close to this may be coming in 8.0).

    These are just a few scenarios in the simplest driving environment--on a highway. While some of these scenarios would clearly benefit from additional hardware, some of them should be perfectly doable with existing hardware: notably 1, 6, 7, 8, 9, and 10. There are more examples of improvements that can come with existing hardware. For example:

    -Stoplight and stop sign recognition.
    -Automatic speed limit adjustment.
    -When following a vehicle, the car should maintain itself over the path that the lead car took over the ground. (This is useful when following a car on a road without lane markings, or when passing through an intersection without lane markings). Instead, it tends to somewhat "cut the corner" and head straight for the lead car, which could put you into the curb or an adjacent lane (or an adjacent car!)
    -If following a car into an unmarked intersection and the lead car changes lanes, your car will follow it into the adjacent lane! This will greatly surprise the car next to you and could lead to a bad day.
    -In stop and go traffic, if the lead car alternates between moving a few feet forward, then stopping...then moving forward, then stopping...then your car will annoyingly accelerate, brake, etc. Car should be able to recognize that the time average speed of the lead car is low, and glide along gradually at a very slow speed, using little energy or brakes and leading to a smooth slow ride.

    These are additional scenarios that are perfectly achievable with existing hardware.

    So over a year of autopilot, we've seen improvement in lane holding, smoothness of steering (and, to a lesser extent, braking). contrast, and other things. But there is a *LONG* way to go before we've exhausted AP1.0 hardware capabilities.

    So while it's possible that, within a few quarters, Tesla could put out cars with hardware for full autonomy (or at least level 3) and then update them via software over time, I wouldn't expect fully autonomous hardware anytime soon. Why?

    1. Achieving level 4 is clearly an interative process. Starting out, Tesla thought they could rely on cameras. After the Joshua Brown accident, they realized the limitation of using cameras as a primary sensor (I believe this caused Tesla to end the relationship with Mobileye) and changed their focus. This will probably happen again. For instance, Elon might be against LIDAR, but he may come around to it if radar processing doesn't turn out to work as well as he'd hoped.

    2. Each sensor that's added makes the processing and software that much more involved and complicated.

    3. Level 3 and certainly 4 will require an enormous amount of testing and validation. By Elon's benchmark, it has to be an order of magnitude safer than a human driver at the very least. Probably several years worth of testing data once they have a Level 4 system before Tesla says you can ride as a passenger while the car drives. But certainly even longer before the government says it's ok..

    4.
    Building a bunch of cars with level 4 hardware and selling them to customers with a promise that they might be able to use them for level 4 driving 5 or more years down the road doesn't make sense. That would be a money-losing proposition for Tesla, unless there are enough foolish buyers out there to pay for the feature many years before it could potentially even be usable.

    Yes, Elon mentioned the car being able to drive from NY to pick you up in LA in about 2 years. (Guess what? As awesome as he is, did anyone ever notice that he's a bit overly optimistic when it comes to time frames?)

    Yes, what Tesla is doing is awesome.


    Yes, Tesla's gathering FAR more data than *anyone* out there.

    But if you extrapolate the improvements we've seen since 7.0 and project that out toward even Level 3 autonomy, you should be able to recognize it will be several years before you can chill out and watch your Tesla automatically merge with traffic and autonomously change lanes. And even longer before it will navigate an intersection and make a turn for you.

    Yes, there are "marketing" and "research" videos out there showing what approaches Level 3 and 4 technology. But while you might see something near Level 3 or 4 autonomy in the video, that's a very narrow subset of the curveballs the world can throw at you.

    I hate to burst any bubbles, but putting a Level 4 car out there in the real world is many orders of magnitude harder.

    Long story short, if you're waiting for the "fully autonomous Tesla" before you put down your deposit, I recommend either buying now or moving on. It's going to be awhile.

    It's going to be interesting as hell.

    But it's going to be awhile.
     
    • x 52
    • x 6
    • x 5
    • x 3
    • x 3
    • x 1
  2. Wknapp0924

    Wknapp0924 Member

    Joined:
    Jan 17, 2016
    Messages:
    154
    Location:
    Oceanside, Ca
    I would like to add just two words to why autonomous vehicles is still at least 5 years away:
    Government Regulation.
     
    • Like x 6
    • Dislike x 2
  3. LargeHamCollider

    LargeHamCollider Battery cells != scalable

    Joined:
    Jan 10, 2015
    Messages:
    477
    Location:
    United States
    You're assuming (or so it appears) that when level 4 comes it will come everywhere all at once. I don't expect this to be the case, I expect level 4 on ramp to off ramp in the next 3 years with the number of "approved routes" expanding from there.
     
  4. ⚡️ELECTROMAN⚡️

    ⚡️ELECTROMAN⚡️ Liked, but not well liked.

    Joined:
    Jul 15, 2016
    Messages:
    346
    Location:
    Oregon
    It's not rocket science to add a few more cameras and RADAR sensors. I don't think the next hardware upgrade will enable full autonomy, but it could be a big improvement over the current hardware. I'll be surprised if they don't start installing more sensors by the end of the year. Elon has recently said that the hardware exists and that full autonomy is coming sooner than you expect. I don't think a genius would say this unless he was planning on upgrading the AP hardware in the very near future.
     
    • Like x 5
    • Dislike x 1
  5. Canuck

    Canuck Active Member

    Joined:
    Nov 30, 2013
    Messages:
    3,959
    Location:
    South Surrey, BC
    #5 Canuck, Sep 12, 2016
    Last edited: Sep 12, 2016
    Yes, you've hit the nail on the head. In my view, the OP jumps from the current system to full autonomy and suggests nothing will take place in the interim. I highly doubt it. I mean, this is Tesla we're talking about. The Model 3 is due out next year. I can't see it coming out with AP1.0 hardware so they need to get 2.0 on the S/X before it's release. Will it be some months away? Possibly. But years away? I doubt it. Will it be full autonomy? Of course not.

    Aside: I sure hate the rating system here since while I disagree with the OP, I more than respect his well thought out opinion so I would never give him a "Dislike". I'm certain there's a lot of people with me on that. So he gets well-deserved "Likes" such that people reading here get a skewed view of things. Why can't we respectfully "Disagree" so we can truly see what people think of his thoughtful post? But I guess our constant calls to do away with this adolescent rating system just falls on deaf ears. I sure miss the old website!
     
    • Like x 2
  6. xav-

    xav- Member

    Joined:
    May 26, 2016
    Messages:
    422
    Location:
    Orange County CA
    So AP20 is quarters away?

    Then why doesn't Elon musk or tesla clarifies that it's not in scope for the near/medium term? There are a lot of folks postponing orders because of this.
     
    • Like x 2
  7. ecarfan

    ecarfan Well-Known Member

    Joined:
    Sep 21, 2013
    Messages:
    10,375
    Location:
    San Mateo, CA
    Todd, I respect your opinion, and appreciate your thoughtful post. But I disagree regarding how soon the next version of Autopilot hardware is coming. I think it is very soon, as in weeks not months.

    AP 2.0 hardware will of course not be capable of Level 4 next year, or the year after, but I think it likely it will be capable of Level 4 within 5 years or maybe even 3 due to Tesla fleet learning and iterative software improvement that Elon is clearly determined to push forward.

    I think once 2.0 comes out and with the fleet learning that is already a in progress, Autopilot is going to improve much more rapidly than it has over the past 12 months.

    We shall see...
     
    • Like x 1
  8. Bet TSLA

    Bet TSLA Member

    Joined:
    Dec 8, 2014
    Messages:
    468
    Location:
    Cupertino, CA
    I don't think it's a good idea to conflate the advent of AP2.0 sensors and the software the implements AP2.0. I think the sensors are coming by the end of the year. Then we'll see a slow ramp of software capability over the next couple of years.
     
    • Like x 2
  9. Cosmacelf

    Cosmacelf Active Member

    Joined:
    Mar 6, 2013
    Messages:
    3,399
    Location:
    San Diego
    I know this is hard to believe, but Elon doesn't spend hours on these forums like we do. I'm pretty sure he is blissfully unaware that some people think V2.0 AP hardware is just around the corner.

    FWIW, I too think AP hardware won't show up until mid year next year at the earliest. Remember that while an upgraded sensor suite is relatively cheap, an upgraded computer and special purpose chips to actually use the new sensors is NOT cheap, nor is it something Tesla can whip up quickly. The Mobileye chipset is very powerful, but the new AP sensor suite will demand probably a 10x increase in compute power over that.
     
    • Like x 2
    • Helpful x 1
  10. Funky

    Funky Member

    Joined:
    Jan 1, 2016
    Messages:
    6
    Location:
    Hong Kong
    I do believe full autonomy is 5+ years away as there are too many corner cases to consider. However, one area I believe the OP may be erroring on is the fallacy around predicting future technology enhancements. Often when you ask someone to make a prediction 10 years into the future on technology, they look at what society has achieved in the last 10 years and extrapolate an equivalent. However, due to the speed of technology innovation, the curve is almost always accelerating so we achieve say the equivalent of the last 10 years in the next 2 years etc. The blog website Wait But Why does a great article on this.

    Therefore, I expect the enhancements to improve significant faster than they have in the past, especially considering the fleet size continues to grow.
     
    • Like x 3
  11. STbreaker

    STbreaker Member

    Joined:
    May 5, 2016
    Messages:
    218
    Location:
    Ann Arbor, Michigan
    I generally agree with the points of the OP, but I think things are going to advance faster than most of us think. I mean, 10 years ago I didn't think I would now be driving a performance all electric sedan that can drive itself on the highway. While Tesla is a business and a publicly shared company, it somehow continues to defy conventional expectations. Obviously people criticize Musk for over-promising and under-delivering, but he continues to push forward and lead the way. Ultimately, it's the boldness of the company that is the difference. I don't doubt that BMW, Mercedes, Porsche could have made a "Model S" first, but they were probably too frightened at the prospect of failing.

    I think we will steadily get more level 3 features within the next two years. Level 4 is hard to predict because I think it requires more buy in from others to be successful. I doubt hardware is going to be the limiting factor in any case - we may not even need much more than we have now even. I'm sure the programmers can come up with algorithms for efficient autonomous driving in ideal circumstances, but people are just too unpredictable. Look at all the negative press autopilot has been subject to this year. Imagine the blame game that would be going on if a "level 4" vehicle hits someone turning left from the right lane - wasn't it supposed to have a contingency plan for that?! So even if Tesla is able to get a car to Level 4 in say 5 years, it probably won't really be plausible for daily use until other cars on the road are also more autonomous. I suppose more insulation might be an alternative solution - maybe a "carpool lane" for level 3 and 4 vehicles?
     
    • Like x 1
  12. Troy

    Troy Member

    Joined:
    Aug 24, 2015
    Messages:
    451
    Location:
    ◘ In THIS video (after 39:35s) Elon basically confirmed that the Model 3 will have autopilot 2.0 hardware.
    ◘ Elon also said, "The Model S and the X are always going to be our technology leader." (source)
    ◘ And we know that the deadline for Model 3 suppliers is 1st July 2017. (source)

    Therefore it is obvious that Model S and X will have autopilot 2.0 hardware before 1 July 2017. My best guess is January 2017.
     
    • Informative x 1
    • Like x 1
  13. 3Victoria

    3Victoria Member

    Joined:
    May 8, 2016
    Messages:
    813
    Location:
    Victoria, British Columbia
    #13 3Victoria, Sep 13, 2016
    Last edited: Sep 13, 2016
    Good thread. The numbering scheme shown have been level 1, 10, 100, 1000 :) rather than 1, 2, 3, 4.

    I also think new sensors will be installed soon, including at least 4 corner radar and 2-3 front cameras. These would allow safer merging and lane changes at least. I would also like complete 360 camera coverage. Price on all of these is falling impressively.

    Processing power will also need to increase massively, but NVidia chips might work well, or their Tesla developed alternatives. I would think an image processor per radar or camera, and additional compute power for integration, logistics and control will be necessary.

    I suspect the radar will incrementally improve such that its resolution achieves to close to that of Lidar, particularly the newer cheaper and solid-state Lidar, and as such Lidar may not be needed since radar has better penetration.

    Humans do pretty well with two cameras, albeit on a swivelling platform (ie our head), so a multi-sensor set-up should be able to do better -- one can image how well we would do with extra eyes in the back of our heads ;-)

    I certainly vote for a (close to) full set of sensors by the Model 3, even if the processing is not available, since that can be a replacible black box. This would at least allow data collection with the new sensors.

    I also agree that we will move faster than most of us predict. I do like the thoight that highway might become level 3/4 before city/rural roads.
     
    • Like x 1
  14. Bet TSLA

    Bet TSLA Member

    Joined:
    Dec 8, 2014
    Messages:
    468
    Location:
    Cupertino, CA
    Change "will have" to "is planned and hoped to have" and I'm with you. Elon's plans often come in beyond their time targets.
     
  15. calisnow

    calisnow Active Member

    Joined:
    Oct 11, 2014
    Messages:
    1,137
    Location:
    Los Angeles
    Todd, what most of us mean by AP 2.0 is simply a next generation hardware suite with multiple cameras and more radars - and a system with vastly improved autonomous capabilities.

    I, for one, am not here to argue about what Level 3 vs Level 4 vs full autonomy means, yada yada.

    So given that definition of AP 2.0 - all evidence (and there is a lot of it) points to December-February timeframe for the hardware to ship in cars.

    This evidence is found in bullet lists on other posts but includes things such as the Model 3 Reveal Part 2 coming at the end of the year. If Tesla shows off a very advanced autopilot suite for Model 3 and does not simultaneously ship that suite in Model S/X they will Osborne their existing fleet. They aren't going to do that and you know it.

    Next generation hardware by end of 2016 beginning of 2017.
     
    • Like x 2
  16. Todd Burch

    Todd Burch Electron Pilot

    Joined:
    Nov 3, 2009
    Messages:
    4,734
    Location:
    Smithfield, VA
    Software is 99% of the challenge. The hardware is relatively easy in comparison.

    As I said, the hardware may be coming within a few quarters. Tesla obviously has the ability to upgrade the software after the fact. But I still believe even onramp to offramp autonomy is still many years away. Those thinking it is coming soon (such as those who think it will be ready around the time the Model 3 comes out) are, in my opinion, being way too optimistic.
     
  17. bob_p

    bob_p Member

    Joined:
    Apr 5, 2012
    Messages:
    939
    AP 2.0 is likely to have improvements in the sensors and more processing power.

    Tesla is surely aware there are potential owners who are waiting until they can purchase the next generation of AP hardware.

    And they also know that if they talk about the AP 2.0 hardware before it is ready to be released, they risk depressing short term sales (which is why they have always announced new hardware features when or after those features have gone into production).

    Because there is likely a sizable number of potential owners waiting for AP 2.0, if they aren't planning to release AP 2.0 any time soon, Tesla may benefit from stating that - something like when Musk stated a while back that "no major changes are planned for the next year".

    We're planning to order a replacement for our classic P85 - and because we expect to keep that car for quite a while, we'll likely wait until we can get at least a 100D battery pack and AP 2.0 hardware, which should provide us about 20% more range - and even if the AP 2.0 hardware still isn't good enough to fully support self-driving, it would likely get closer than the AP 1.0 hardware.
     
    • Like x 1
  18. HankLloydRight

    HankLloydRight Fluxing

    Joined:
    Jan 18, 2014
    Messages:
    5,785
    Location:
    Connecticut
    I am in total agreement with the OP, and I would say "full autonomy" is even further out.

    They've also been promising us "AI" since the 80's and where is that today? Nowhere really unless you're talking about playing chess or playing Jeopardy. AP is really not much different than AI.

    As I've said before, full autonomy won't really be possible until you completely eliminate the unpredictable human factors on the road. With more and more distracted drivers using their phones, more and more cars will be doing unpredictable (even for Tesla level AP software) things.

    Or even things that are easy for humans to see and avoid in the road, like truck tire shreds, could prove very difficult for radar or cameras to pick up and SAFELY swerve to avoid. If you've ever run over one of those things, it could range from uneventful to a deadly accident.

    For instance, what if there is a huge tire shred in one or even two lanes away from you and not right in front where sensors might pick it up? You can see it, you know that car in front of you is about to swerve wildly to avoid it, and you also move out of the way to give that car space to avoid the obstacle. No way AP could predict, handle, and safely maneuver around an obstacle in some other lane.

    I think there should be a "Turing Test" for autonomous driving, maybe call it the "Musk Test", where there are passengers in a car driving down the road and the "driver" is hidden. If the passengers can't tell if the driver is human or a computer, it passes. I don't think we'll get there in at least 10 years, if not longer. And that's NOT taking government regulations into account.
     
    • Like x 5
  19. Dynastar

    Dynastar Member

    Joined:
    Dec 7, 2015
    Messages:
    107
    Location:
    Billings, MT
    I don't know how long level 4 will take, but I disagree with how long new hardware will take. For one, we've had several reports from fairly reliable sources that the change was coming. Some may not agree, but they've been reliable in the past.

    In addition Tesla needs to keep a tech edge over its rivals. AP is one of Tesla's "killer aps". As it is Tesla's blind spot detection is pretty terrible, which the rumored corner radar would fix. Even if AP 2 software isn't ready that's never stopped Tesla before- how long did we wait for AP 1 to be fully switched on?
     
  20. u00mem9

    u00mem9 Member

    Joined:
    Jun 8, 2016
    Messages:
    289
    Location:
    USA
    Agree with the OP, level 4 is much further off than Elon and co imagine.

    But, there will be AP2.0, AP3.0, etc. because they will want to increase capability as they realize Level 4 is far away.

    Also, just so all are clear, Level 4 is defined by the industry / regulators and essentially means the car drives point to point without intervention AND (this is the huge part, that everyone including Tesla seems to marginalize when making the strategy) if anything goes wrong with the vehicle and it's actions, Tesla is going to be liable.

    Elon is guilty of thinking about this as a purely technical challenge. There is a HUGE cost to defending yourself and the company at Level 4 even if the system works perfectly and is never at fault. Every accident is initially going to result in a lawsuit that Tesla will require defense against at great cost.
     
    • Helpful x 1
    • Like x 1

Share This Page