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

about to buy a cpo model s. IS AP2 that much better than AP1?

This site may earn commission on affiliate links.
So for the record: AP1 is not obsolete.
It’s doing a better job at doing what AP2 is supposed to do well. Still today: lane keeping, lane change and tacc are butter smooth with AP1.

That being said, AP2 is able to do stuff AP1 was never designed to be able to (eg NoA, advanced summon, FSD).

So if you need more that what AP1 can do, definitely go for AP2.
 
Ap1 was designed by Mobileye...the leader in autonomous systems. Tesla and Mobileye had a dispute and parted ways in 2016. tesla was forced to develop their own in-house AP and thought they could do a much better job. Bottom line, AP1 worked batter than AP2 with half the cameras, 1/4 the processing capability. AP2 wasn’t double the price for half the functionality initially. It’s taken now 3 years for AP2 to catch up to its predecessor and still lags in certain areas.

If the dispute had not occurred, FSD would probably have been in full use by now.

Tesla was claiming more functionality and that AP was like hands free driving. Mobileye kept warning them about the limitations on the software and its abilities. Dispute occured after the first major AP accident killing the driver. Meye has enough and as a condition of leaving, Tesla couldn’t no longer sell their patented AP. Hence the birth of AP2...
 
Not sure I agree with your summary or assessment of how the relationship broke down, but everyone is entitled to their opinion and both Tesla and Mobileye have told their (different) sides of the story publicly.

For those who say that AP 2 (or 2.5) are not better than AP1, that's simply not true anymore. The ONE thing that AP1 does that AP 2 and 2.5 currently do NOT do is read speed limit signs. This is not because the can't...it's because Mobileye was given a patent for reading speed limit signs by using image recognition techniques that are pretty much the ONLY way to recognize and read a speed (or any other traffic sign). The result is that they are threatening to sue anyone else who offers this for patent infringement and no one has fought them on it in court. Tesla (and others) have been unwilling to pay Mobileye's desired license fee to use their "proprietary" technique and so at least for no, AP2 and 2.5 cars don't have that feature.

I won't get into a big debate about the validity of Mobileye's patents. Like many other companies, they are patenting all kinds of stuff that probably should not be granted patents (kind of like the competing Samsung/Apple Patents over a rectangular phone with a glass screen and curved corners), but that's not the point of this thread.

Having driven both, I personally find that the current versions of AP 2+ software to be far superior than AP1 software for smoothness, situational awareness and automated driving (especially NOA on highways). Others aren't satisfied with it (and probably never will be) but again, everyone is entitled to their own opinion.

If I was buying a CPO Tesla today, I would personally stay away from AP1 cars because they are effectively hitting end of upgrades. Doesn't mean they aren't great and capable cars, but if you own cars as long as I do (8-10 years), you'll just watch over the next 8 years as more and more features come out you have no hope of receiving (from an AP perspective). That's just my opinion though.
 
AP1 is technically already obsolete. It’s reached its end of life and won’t be updated in the future. Having said that, you just need to decide if it’s worth the price difference to you. When I bought my used Tesla last year, AP2 cars were much more expensive and after reading up on the features I decided it wasn’t worth it. I drive in traffic on the freeway everyday and AP1 works great for that. I don’t need NOA, summon, or sentry mode, although dashcam would be a nice bonus.

I also tend to only keep cars for a few years at most.
 
Not sure I agree with your summary or assessment of how the relationship broke down, but everyone is entitled to their opinion and both Tesla and Mobileye have told their (different) sides of the story publicly.

For those who say that AP 2 (or 2.5) are not better than AP1, that's simply not true anymore. The ONE thing that AP1 does that AP 2 and 2.5 currently do NOT do is read speed limit signs. This is not because the can't...it's because Mobileye was given a patent for reading speed limit signs by using image recognition techniques that are pretty much the ONLY way to recognize and read a speed (or any other traffic sign). The result is that they are threatening to sue anyone else who offers this for patent infringement and no one has fought them on it in court. Tesla (and others) have been unwilling to pay Mobileye's desired license fee to use their "proprietary" technique and so at least for no, AP2 and 2.5 cars don't have that feature.

I won't get into a big debate about the validity of Mobileye's patents. Like many other companies, they are patenting all kinds of stuff that probably should not be granted patents (kind of like the competing Samsung/Apple Patents over a rectangular phone with a glass screen and curved corners), but that's not the point of this thread.

Having driven both, I personally find that the current versions of AP 2+ software to be far superior than AP1 software for smoothness, situational awareness and automated driving (especially NOA on highways). Others aren't satisfied with it (and probably never will be) but again, everyone is entitled to their own opinion.

If I was buying a CPO Tesla today, I would personally stay away from AP1 cars because they are effectively hitting end of upgrades. Doesn't mean they aren't great and capable cars, but if you own cars as long as I do (8-10 years), you'll just watch over the next 8 years as more and more features come out you have no hope of receiving (from an AP perspective). That's just my opinion though.

That's a great summary and the only thing I would add to that is that Mobile Eye is now owned by Intel.

I seriously doubt Intel will take the step of suing Tesla or allow their company MobileEye to sue Tesla, especially given the fact that I am sure by now Tesla has their own portfolio of FSD and other patents that I am sure Intel/Mobile Eye might be encroaching on. I wish some bogus patents were never issued but a patent being issued and it being enforceable are two different things.

In fact it is not unusual for issued patents to be quashed in court when enforcement of that patent is attempted.

Tesla should go for it with speed limit detection by camera. Speed limits change around where I live all the time. One major road just had the limit reduced from 45 to 35 and I really wish the speed limit detection was by camera and not GPS.
 
  • Like
Reactions: drklain
Honestly, no, it isn't. It was initially something I wanted and it made it hard to find a used Model S within the price range I was willing to pay for a used P90D since so few had AP2. After doing more research I determined that AP2 really isn't worth the extra money as I don't believe the hardware in the model S will ever be truly self driving where the user doesn't need to pay attention. There's too little/no redundancy and the software is many years away (probably a decade in my opinion). With that worked out, I figured everything that AP2 offers over AP1 is either pretty minimal or a parlor trick, so I settled on buying AP1 and saving some cash.

Then the big end of quarter sale happened last month and I ended up with a new P100D with AP2.5 and I can safely say that at this point, AP2.5 is 100% not worth spending extra money for over AP1. All that's left is for you to determine if you believe the extra sensors in AP2 are worth your money, if you think that FSD will happen with them it at least some features you think are worth it. For me, I wouldn't spend more than maybe 3K more, 5K tops for it. (Very happy with my car, just not impressed with AP2.5 at all)

Oh also consider the differences in AP2 and AP2.5 and consider that only ~75K AP2 cars even exist, with a subset of those even paid for FSD. I have little confidence in Tesla ever doing FSD on 2.5, and even less on them bothering to support AP2 cars.
 
Honestly, no, it isn't. It was initially something I wanted and it made it hard to find a used Model S within the price range I was willing to pay for a used P90D since so few had AP2. After doing more research I determined that AP2 really isn't worth the extra money as I don't believe the hardware in the model S will ever be truly self driving where the user doesn't need to pay attention. There's too little/no redundancy and the software is many years away (probably a decade in my opinion). With that worked out, I figured everything that AP2 offers over AP1 is either pretty minimal or a parlor trick, so I settled on buying AP1 and saving some cash.

Then the big end of quarter sale happened last month and I ended up with a new P100D with AP2.5 and I can safely say that at this point, AP2.5 is 100% not worth spending extra money for over AP1. All that's left is for you to determine if you believe the extra sensors in AP2 are worth your money, if you think that FSD will happen with them it at least some features you think are worth it. For me, I wouldn't spend more than maybe 3K more, 5K tops for it. (Very happy with my car, just not impressed with AP2.5 at all)

Oh also consider the differences in AP2 and AP2.5 and consider that only ~75K AP2 cars even exist, with a subset of those even paid for FSD. I have little confidence in Tesla ever doing FSD on 2.5, and even less on them bothering to support AP2 cars.

Aren't all the Model 3s they have sold on AP 2.5?
 
Not sure I agree with your summary or assessment of how the relationship broke down, but everyone is entitled to their opinion and both Tesla and Mobileye have told their (different) sides of the story publicly.

I had 2 friends working at both MBeye and Tesla Corp at that time. Both layed down the reality of it and it was quite similar. What was publically portrayed was...well it is what it is :)

As for AP2.5 being better than AP1, yes it is for certain situations. Ie. maintaining lines above a hill. However I remember AP1 through 2016/2017 with our 90D and it was smooth as butter. Held centre of the lane constantly, never phantom braked, smoother deceleration...it just never felt ‘drunk’. AP2 still gives a certain level of mistrust when a phantom brake engages or the car sees a vehicle or object falsely, at times does not hold the Center of the lane and till now detects the wrong speed signs.

This is why I still maintain AP1 was a superior designed product able to perform with higher accuracy and comfort with 4 cameras (not 8) and a lower processing capability. Those of us who have gone through the early stages of AP2 to now can probably relate.

Fast forward to today...unless one is beta testing FSD, the current AP2/2.5 is not that far off AP1 at its current level. If one is happy with semi-autonomous driving only, then AP1 is in my view is still a smoother drive.
 
So as you can see from above, the AP1/AP2 debate is extremely subjective. If you are buying a CPO that you plan to keep for 2-4 years, then I think at this point the differences are minimal. Neither system is capable of FSD (and AP2 probably won't be in the next two years) and have to be considered driver assist applications. It comes down to how much "assistance" you are looking for. AP1 will not receive any enhancements but in its current configuration (including the ability to read speed limit signs) is more than adequate for most folks. I use it all the time, both on highways and back roads) and find it very capable. I use simple summon all the time to get into and out of my garage, autopark for parallel parking works well and simple summon also is handy if there are parking spots with puddles.

I keep my cars for about 3 years and when I am ready to upgrade my S in two years, I will certainly spend the additional money to get AP2 at that point. Right now, I haven't found any compelling reason for AP2.
 
This is how I see it. We have a Model S with AP1, and a Model 3 with AP2.5. AP1 seems like Confident, yet Naive mindset. It handles many AP situations very well and Lane changes are without hesitation. But there are times, where it doesn't have enough information to make a move, "Safely". I has one camera and then only the sensors and if something is just out of range, but is coming in hot, AP1 will still do the lane change without any real concern of what's coming up behind it in the new lane.

AP 2.5 is like a Responsible, yet Timid mindset. It has 8 cameras and lots of sensors, it essentially has information overload and has to parse through all of that to make an informed decision. So there are times when it sees a shadow on the road and freaks out or waits for traffic to be relatively perfect before tip toeing into a Lane Change.

What I'd like to know is how AP1 vs AP2+ learn from their driving experiences with their user. Do they do it the same, do they do it differently? To my recollection, AP1 was not Tesla's project, it was a 3rd Party that partnered with Tesla to create AP1. AP2+ is all Tesla, so that makes me wonder how Data Collection and Learning differ, if at all, between the two.