Tesla FSD - A Five Year Perspective

My Model 3 was delivered in mid-August 2018. While I originally just had Enhanced Autopilot (EAP), I quickly paid the extra $2000 to get FSD as rumors swirled about Tesla charging more in the future. What's my opinion after 5 years of testing and playing with Tesla EAP and FSD? TLDR: It's pretty much still just a gimmick and hardly worth the extra $2K I paid for FSD, much less the $15,000 they charge today.

So a little about my experience so you understand my evaluation. Since receiving the car in August of 2018, I have done everything possible to have the latest firmware, latest features, and test and provide feedback on the latest software and capabilities. I am an engineer by training and education, and have participated in a lot of software Beta Test programs for lots of products, and I take the whole thing fairly seriously because to me it's lots of fun. So I don't feel like I have (or had) unrealistic expectations of what FSD would be able to do, what the testing process would be like, or when the features would become polished and truly usable.

All that said, let's start with what works: Navigate on Autopilot (NoA) and Auto Lane Change. NoA is an L2 ADAS feature that combines TACC and Autosteer to drive the car on limited access highways, and includes features like automatic speed-based lane changes and highway interchange transitions. IIRC, the initial version was released in October of 2018 and, since then I have put thousand of miles on it. Now a part of FSD-beta on highways, while it has improved somewhat, it basically the same functionality as NoA on release five years ago. Phantom braking is noticeably better (happens less frequently but still happens) and the system now has some subtle safety improvements, like cheating to the outside of the lane when passing a truck. But passing behavior is still wonky and can be harrowing, and lane selection for passing and exit/transition ramps is still really bad - maybe even a bit worse than in original NoA now that it doesn't use maps as much. In addition, when it came out in 2018 I felt like NoA was "best-of-breed" for these L2 ADAS highway-driving systems, but I feel like it has now been surpassed by BlueCruise, SuperCruise, Drive Pilot, and the like, many of which offer some level of hands-off and/or eyes-off driving for a truly L3 autonomous driving experience. But NoA is still a very useful feature, IMO, offering reduced workload, fatigue, and stress on long highway trips, especially if you do your own passing with Auto Lane Change instead of the automated speed-based lane changes.

Another thing that works is Summon. Remote control your car into and out of a garage or tight parking space. Only useful in very limited circumstances but it does what it does well when you need it (as long as you still have USS, evidently).

Everything else: AutoPark, Smart Summon, Autosteer on City Streets, and Traffic and Stop Sign Control? Gimmicks. Barely usable and hardly ever useful. Setting AutoPark and Smart Summon aside (because they just plain don't work), Autosteer on City Streets is a massive undertaking and has improved immensely since "beta" testing began almost three years ago now. But to engage it, you have to be ready to be extremely alert and ever vigilant - hardly a relaxing way to drive. And while it can get you from point A to point B - even sometimes without intervention (which has never happened to me) - it is most often still a harrowing experience to use it. As an ADAS feature, it has zero utility, IMO. Fun to show off to your friends, and fun to test the latest and greatest to see what it can do. But from a practical standpoint, it can't get me anywhere I need to go faster, more efficiently, more relaxing, or, regardless of how many people quote statistics to try and prove otherwise, safer than I can get there by driving myself.

And now looking back 5 years and thinking where we've been and how far we've come, it's actually quite sad to me that this is all we got. I don't know how much more appetite for "testing" FSD beta I have, and often think I would be much better off or newer firmware just driving myself or on plain, old AutoPilot (I, personally, would still have my NoA in my EAP). But if there's anybody out there thinking whether they want to invest $15,000 in Tesla's autonomous driving systems, just take a tip from me and go into it with (very) low expectations of what you will be getting.

 

[Dewg]

But from a practical standpoint, it can't get me anywhere I need to go faster, more efficiently, more relaxing, or, regardless of how many people quote statistics to try and prove otherwise, safer than I can get there by driving myself.

And it never will. Even Waymo, the leader in L4 AVs, and bristling with an advanced sensor package, can't do it. The only thing it can possible do is satisfy your "relaxing" criteria. AVs and L2 ADAS systems will never be faster or more efficient than you driving manually. Safety is debatable. There are people who claim FSD Beta saved them from something they didn't see, and there are people who claim FSD Beta would have caused an accident if not for their intervention. Even the more advanced AVs like Cruise are crashing into buildings, and they have LIDAR, RADAR, USS, and cameras all over them.

 

[diplomat33]

Even Waymo, the leader in L4 AVs, and bristling with an advanced sensor package, can't do it.

I don't think it is fair to say that they can't do it. Maybe not yet but we can't say never. Waymo might be slower than a human driver but it is already safer than a human driver now. Who is to say that Waymo won't eventually achieve both, faster, more efficient and safer.

 

[Dewg]

I don't think it is fair to say that they can't do it. Maybe not yet but we can't say never. Waymo might be slower than a human driver but it is already safer than a human driver now. Who is to say that Waymo won't eventually achieve both, faster, more efficient and safer.

Safety I will absolutely agree with. Faster and more efficient, never. And that's down to 1 easy reason: humans break the law. Waymo will eventually drive faster, I agree, but will never go above the speed limit. I'm late for a meeting, so I'm going to drive 55 or 60MPH in a 45MPH zone. I will roll my stops, and likely turn on red. Waymo, or any of the AVs, will not.

That said, there's no arguing the convenience of AVs. L3+ systems allow me to work on something else while the driving is happening. But they won't be faster than a human driving.

 

[diplomat33]

Safety I will absolutely agree with. Faster and more efficient, never. And that's down to 1 easy reason: humans break the law. Waymo will eventually drive faster, I agree, but will never go above the speed limit. I'm late for a meeting, so I'm going to drive 55 or 60MPH in a 45MPH zone. I will roll my stops, and likely turn on red. Waymo, or any of the AVs, will not.

That said, there's no arguing the convenience of AVs. L3+ systems allow me to work on something else while the driving is happening. But they won't be faster than a human driving.

Maybe not faster than a human but AVs will get more efficient in their routing IMO.

 

[Darren F]

Imagine posting this weeks after the FSD 12 demo. Not to mention it's now $12k, not $15k.

 

[mgs333]

Imagine posting this weeks after the FSD 12 demo. Not to mention it's now $12k, not $15k.

Did you watch the same FSD12 "demo" I did? How would that "demo" drive affect OP's opinion? It largely drove similarly, including trying to go on a red.

 

[Darren F]

Did you watch the same FSD12 "demo" I did? How would that "demo" drive affect OP's opinion? It largely drove similarly, including trying to go on a red.

That software definitely did not perform the same as today's software. It was much smoother, and responding to stimuli that todays software does not. Speed bumps for example. The big takeaway though is the way it functions, not how it functioned. Tesla has been coding FSD ever since MobileEye went away. FSD 12 has none of that, it simply "learned" how to drive, in 2 months. Just last week Tesla turned on a new training compute system that multiplied their previous capability by a factor of nine times. That means that two months training could easily happen in less than 2 weeks now. That is why a post like this at this late stage is daft at best. Obviously the OP's opinion wouldn't be changed by the video, he's already made up his mind. But he's an engineer, so apparently he knows things.

 

[mjwojtukiewicz]

Ok, so V12 is a marvel. Why are we still stuck with V11 or less? How much more dangerous can it be. Ad BTW that demo drive was a walk in the park compared to what alot of urban and suburban drivers encounter

 

[JB47394]

Ok, so V12 is a marvel.

Full stop, right there. V12 is not a marvel. It's an alpha test version of some new software that shows some interesting capabilities. That's it. Everything else is speculation. V12 is so preliminary that we shouldn't even be thinking about it. The Tesla engineers have just started fooling with it, so they don't have any serious experience with it and don't know what kind of legs this approach has. They thought they could make V10 work. They thought they could make V11 work. Now they think they can make V12 work.

 

[kabin]

Ok, so V12 is a marvel. Why are we still stuck with V11 or less? How much more dangerous can it be. Ad BTW that demo drive was a walk in the park compared to what alot of urban and suburban drivers encounter

Agreed. V12 is probably still in early stages of development but after all these years that drive was Elon choreographed. Even then I recall an attempted red light run and roundabout difficulties. Much of the status quo along with Elon overstating v12 performance and how bright the future is. V12 will have more NN implementation but not full, honest to goodness, end to end NN.

 

[Goose66]

That software definitely did not perform the same as today's software. It was much smoother, and responding to stimuli that todays software does not.

My opinion, as made clear in the original post, is expressed based on my own experience with FSD feature set over the last 5 years. This is bound to be different from the opinions of people formed from viewing other people's experience on YouTube. Elon's demo of v12 was just his (one) experience. I don't understand how, if you weren't riding in the vehicle at the time and hadn't previously driven that route in that vehicle on FSD v11.x, you can say with such certainty that it was "much smoother" and "definitely" performed differently than today's software.

That is why a post like this at this late stage is daft at best. Obviously the OP's opinion wouldn't be changed by the video, he's already made up his mind.

Again, my post is based on my personal experience over the last 5 years. Yes, I have made up my mind that the features in FSD, outside of NoA, have been and are now gimmicky and not useful - but nothing more than that. I did not express any opinion on how successful the next versions of the product might be, where things are going with Tesla FSD or other systems, how useful generally L3 autonomous vehicles will be in the future, or any other topics. I could certainly debate all of those with you, but none of that was said or implied here.

But he's an engineer, so apparently he knows things.

Yep. I know things and I fix stuff. That's the essence of being an engineer.

 

[kabin]

Full stop, right there. V12 is not a marvel. It's an alpha test version of some new software that shows some interesting capabilities. That's it. Everything else is speculation. V12 is so preliminary that we shouldn't even be thinking about it. The Tesla engineers have just started fooling with it, so they don't have any serious experience with it and don't know what kind of legs this approach has. They thought they could make V10 work. They thought they could make V11 work. Now they think they can make V12 work.

I think you two were saying the same thing in different ways.

 

[GalacticHero]

My experience mirrors this very closely. I'm also an engineer (software, mostly firmware and recently machine learning and computer vision.) I bought my Model 3 in November 2018 with EAP and added FSD with the same $2,000 deal. I'll agree that NoA works, once one turns off lane changes for passing and forgive the phantom braking. It has not done any of the "slammer" brakes for a while, but regularly slows down abruptly when going under certain overpasses. Regular fwd-bwd Summon works for squeezing in and out of tight parking, but it not useful for other parking adjustments since the time for the app to link up is so slow I could easily back into and out of the car quicker. [Otherwise it would be useful for times I get out and decide I should have been better centered in the space.]

I'm a little more lenient on the FSD in city. I use it on the non turning portions of drives and find it does a good job at recognizing stop lights and stop signs. However, unless there are no other cars around and I'm bored, I always disengage it before turns because it quite often does something unsafe, and will always be maddeningly slow. It also still has the problem with wanting to make wrong lane changes, including failing to make necessary lane changes.

Basic TACC also has a problem. It usually quits working during precipitation, either due to the visibility or cameras getting blocked. It really really needs to fallback to standard cruise control in those cases.

 

[Padelford]

Ok, so V12 is a marvel. Why are we still stuck with V11 or less? How much more dangerous can it be. Ad BTW that demo drive was a walk in the park compared to what alot of urban and suburban drivers encounter

I too have been in this process for at least 6 years, and have become very skeptical and disappointed with what we've been given vs promised in Autopilot and FSD (let alone Autopark and Summon)

Let's get real for a moment. Consider the history of Autopilot and FSD, especially all the false assertions made about upcoming capabilities. Recall the 2016 video that supposedly showed what Autopilot software was shortly going to be capable of.

Why should we think the "V12" demonstration was really what Musk claimed? He never showed the Software page on the vehicle to demonstrate that he was indeed running V12 software, although that could have been faked as well. I noted that v12 made some of the same mistakes as V11 does now, including making a left turn into the wrong lane. OTOH, based on my experience, V11 wouldn't have tried to run the red light like V12 did.
I think everyone should really temper their expectations of what V12 will bring. It's much more likely to be a disappointment than a great stride forward.

 

[Supcom]

Why should we think the "V12" demonstration was really what Musk claimed? He never showed the Software page on the vehicle to demonstrate that he was indeed running V12 software, although that could have been faked as well.

Maybe there was a guy in the back seat using a game controller to maneuver the car. Elon kept blocking his view, leading to driving errors like missing the red light.

If you're going to get all conspiratorial, at least do it right.

 

[EvilNuff]

What absolutely shocks the hell out of me is that Tesla is even *talking* robotaxi. (And apparently Elon is still pushing for a controller less car for some insane reason...) In order to accommodate a robotaxi scenario FSD will have to be solved to 100%. It has to account for everything it could encounter. As anyone in software can tell you solving a software problem to 100% is an exponential difficulty curve. If they are currently at 80% (being extremely generous there) towards level 5 autonomy that might not even be half way to a robotaxi solution. Folks we are decades away from robotaxis with Tesla's chosen solution. The race isn't over but the horses are coming down the final stretch and Tesla is dozens of lengths behind Waymo at this point.

 

[dpeilow]

What's worse is hobbling your existing cars by removing driving controls because you are so convinced this world of autonomous driving is just around the corner.

 

[zroger73]

In order to accommodate a robotaxi scenario FSD will have to be solved to 100%.

While 100% would be ideal, humans don't and can't drive 100% - they have wrecks and get tickets around the clock.

 

[zroger73]

...including trying to go on a red.

Only one red light run in a 45-minute drive is impressive to me as someone who only has experience with FDSb 11.3.6 and 11.4.4.