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

Does anyone else have zero interest in EAP?

This site may earn commission on affiliate links.
Paying attention is a requirement to driving, AP or not.

Yes, but paying attention to the necessary degree is natural when driving and completely unnatural when being driven. With a little bit of exaggeration, it’s like sensing the trajectory of an incoming tennis ball and hitting it back just right versus sitting down and doing the math to calculate the incoming trajectory, speed, desired point of impact etc.

That being said, its not for everyone. Some people aren't ready to give up control of driving to a computer. Wondering constantly if the car will behave or kill you wouldn't be a good experience.

I don’t think it’s so much being unwilling to give up control but rather being willing to hand control to an insufficiently capable system.

I’ve beed observing Autopilot behavior quite keenly on the various loaners I’ve been driving while my 3 is in the shop for the customary remedial manufacturing activities.
Objectively, it drives poorly (judging simply by the frequncy of its murder attempts). Subjectively, it’s also largely unpleasant and i find driving with TACC alone to be a lot more relaxing. Also, now that I’m driving an older AP1 Model S loaner, I can state that unfortunately AP1 is still better at the basics of driving. There may be some things that AP2 does better but, at the core, AP1 is still a superior system (probably solely due to its ability to more accurately measure relative speed using the front radar).
 
  • Like
Reactions: V__2
I think it boils down to:

If you drive in stop and go traffic it's a must have.
If you do long drives with a wide open freeway it's nice.
Otherwise it can be nice, but it depends on how much you like it (personally I use it all the time in the first case, sometimes in the second and never at speed when there's a lot of cars around or tight shoulders).
Fair assessment. Personally I'm not a fan of AP on surface streets, even if it helps with stop-and-go traffic. Freeway driving (read: long trip), it's a must-have. Sooooo handy. Overall I'm happy with the $5k purchase, and I'm glad it's improving. Looking forward to the advanced summon features Elon's been talking about.
 
I think it boils down to:

If you drive in stop and go traffic it's a must have.
If you do long drives with a wide open freeway it's nice.
Otherwise it can be nice, but it depends on how much you like it
(personally I use it all the time in the first case, sometimes in the second
and never at speed when there's a lot of cars around or tight shoulders).
This perfectly describes the typical use case.

EAP also actually distinguish Tesla from any other cars.

In the particular the V9 2018.42 version provides some noticeable enhancements.
There are a lot of YouTube videos dealing with this subject.

IMO, unless you really want to save some dough, which comparing the overall cost is not that much,
a Tesla without EAP is a kind of 'no-no'.

When you sit in an EV, you feel that the Kodak moment had already occurred for the other ICE cars.

With EAP, you feel the SpaceX moment.
 
Last edited:
  • Like
Reactions: MaryAnning3
Yes, but paying attention to the necessary degree is natural when driving and completely unnatural when being driven. With a little bit of exaggeration, it’s like sensing the trajectory of an incoming tennis ball and hitting it back just right versus sitting down and doing the math to calculate the incoming trajectory, speed, desired point of impact etc.

I don’t think it’s so much being unwilling to give up control but rather being willing to hand control to an insufficiently capable system.

I’ve been observing Autopilot behavior quite keenly on the various loaners I’ve been driving while my 3 is in the shop for the customary remedial manufacturing activities.
Objectively, it drives poorly (judging simply by the frequncy of its murder attempts). Subjectively, it’s also largely unpleasant and i find driving with TACC alone to be a lot more relaxing. Also, now that I’m driving an older AP1 Model S loaner, I can state that unfortunately AP1 is still better at the basics of driving. There may be some things that AP2 does better but, at the core, AP1 is still a superior system (probably solely due to its ability to more accurately measure relative speed using the front radar).

I can see where you are coming from, just doesn't occur like that to me. Since driving is second nature to most of us by now, being driven by ap is more like putting your hands on the wheel while another driver ( I imagine a 16 y/o teenager) is doing the same on the same wheel. As long as the muscles of both of you agree, its still pretty automatic. You feel the wheel moving for you and its not until it does something you object to that your attention starts yelling at you. If you treat the car like a 16 y/o driver its driving skill seems appropriate.

Since I only use it on divided freeways, to me it is very very very good at that driving. I am sure that with my being in the Bay Area, there is much better data for the local freeways. There are a few things I wish it did better, but overall I find it great at doing the long boring stretches with or without stop and go. I send bug reports every single time it does something objectionable even simple things as I hope there is someone looking at this data.

I have yet to encounter any "Murder Events" or any event where the car failed so badly that imminent deadly collision was impending. The worst near collision I have experienced is when merging at 35 mph, it simply wouldn't give the other driver room to merge, as it was following the speed of the car ahead of me in my lane. I took control when the gap between my front fender and their rear fender was 18" I do not know if it would have recognized a collision impending at 12" or just run into the car next to me.

Where it needs a lot of work:
Merging- Car basically ignores other vehicle blinkers, and other cars half in your lane but moving in until it recognizes them as an object in your path and brakes suddenly. Wish it would look a bit at the cars directly in front and behind of you in the adjacent lanes and make better decisions.
Changing lanes - If there is a spot open in traffic flow right next to you and you trigger a lane change, its first choice is to slow down, even if maintaining speed or speeding up slightly would complete the lane change.
Phantom braking - It really likes to brake excessively when the car in front of you changes lanes out of your lane. I assume this behavior is to give maximum response time to the AP system in case the car in front of you is dodging an object. It would be very difficult to stop in time for a pallet or something like that in the fast lane if the car in front of you barely could avoid it by changing lanes.
Display - Currently other vehicles are displayed in a very jumpy manner, not very confidence inspiring that the computer really has a handle on all of the traffic
General awareness of traffic - related to merging, but in essence the car doesn't seem to look beyond the cars immediately adjacent to it. If it knew that a merge was coming such as an off or on ramp it could perhaps give a bit of space, and try to make the "Zipper" work as merging

Things I haven't yet experienced but want to, to give me a good idea as to the capabilities of the system, and increase my knowledge of where to trust it:
Automatic evasion from an incoming car about to crash into you.
Automatic stopping or avoidance due to debris/obstructions in the road. I don't even know what it would take to trigger an emergency stop event. Would a pallet be enough? How about a tire or step ladder?
Avoidance or warning of a pedestrian or bicycle who was driving poorly or where they weren't supposed to
 
I can see where you are coming from, just doesn't occur like that to me. Since driving is second nature to most of us by now, being driven by ap is more like putting your hands on the wheel while another driver ( I imagine a 16 y/o teenager) is doing the same on the same wheel. As long as the muscles of both of you agree, its still pretty automatic. You feel the wheel moving for you and its not until it does something you object to that your attention starts yelling at you. If you treat the car like a 16 y/o driver its driving skill seems appropriate.

That’s my feeling exactly, although I’d say 16 y/o, nearsighted and not wearing their glasses. It’s making decisions based on the immediate vicinity, kind of like playing chess while thinking one single move ahead. No matter, if I were still doing the 680 crawl, I would have purchased it.


Everything I have experienced and read about Tesla’s Autopilot software leads me to believe two things:

1. The software infrastructure is nowhere near as sophisticated as people may think.

I hear Elon saying that everything should be feasible with just vision. Technically that’s true, but it’s much harder and failure prone. Performing quick image recognition with some crummy neural network will be a lot less effective than doing the same augmented with the spatial awareness provided by an array of radar/lidar sensors. Just look at how long it took to release the auto-wipers and how poorly they perform at night or in light rain or fog.

2. Until recently, I don’t think Tesla dedicated much effort to Autopilot improvement. Now that some of the Model 3 hoopla is behind them, more notable upgrades seem to emerge more frequently.

To me it appears that Tesla performs in an excessively single-threaded fashion, despite its rather substantial size. While that was totally understandable and likely necessary in the early days, I think it’s a big hindrance at this point. This is probably due to organizational issues (centralism with a sprinkling of cultism) and will probably need a crash event of some sort before it changed.
 
Fair assessment. Personally I'm not a fan of AP on surface streets, even if it helps with stop-and-go traffic. Freeway driving (read: long trip), it's a must-have. Sooooo handy. Overall I'm happy with the $5k purchase, and I'm glad it's improving. Looking forward to the advanced summon features Elon's been talking about.

Yeah, I'm talking about stop and go freeway driving. I never use EAP on surface streets.
 
I'm surprised that I keep hearing that EAP is a small percentage of the cost.
Most of us are getting the $7500 rebate and thus the price of higher cost perf version is 71K. Subtract $7500 and that becomes $63,500.
Thus $5k or $5.5k is ~8.7% (not an insignificant percentage).
For a RWD or Mid, it becomes something like 13%
 
I think the features are nice. But, yeah, that extra cost wasn't worth it to me. Especially considering how I like to feel more in control. Granted, the standard features do help me in some instances, no doubt.
 
I didn't get EAP because I didn't think it would be worth the money, and I was willing to pay the extra price later if I decided I really wanted it.

Then I got the trial and it only made me want it less. It's way too nerve racking to use at speed. Nice in traffic, I guess, but I still find myself second guessing it, wishing it was slightly more aggressive sometimes, less aggressive other times. I guess I'm a control freak. It's just not appealing.

I recently went from bike commuting to commuting 45mi/day in the 3, though my wife and I share the driving. Traffic flows very quickly for the most part, so no stop and go grind.
 
The way I look at it is this: my product dev/qa background makes me pay close attention to what works and what doesn’t. I find that I can guess pretty well what has a good chance of working well (such as low-speed traffic with lots of reference points around) and conversely what contexts the car is likely to screw up in. I wouldn’t pay money for AP, but I could find it entertaining in certain contexts.

I don’t think other family members would be keenly-enough aware of the shortcomings and will guaranteed end up in some dangerous oh-*sugar* situations. I’m not willing to gamble that.
 
I think on a major freeway (I-5, I-84, etc...) the EAP is pretty awesome and IMO it probably will save you more often that it will cause problems specially driving at night. I definitely think it could be a lifesaver if you fell asleep at the wheel or something. Whether it is worth the extra money pretty much depends on how many freeway miles you log, for now anyway.
 
I think on a major freeway (I-5, I-84, etc...) the EAP is pretty awesome and IMO it probably will save you more often that it will cause problems specially driving at night. I definitely think it could be a lifesaver if you fell asleep at the wheel or something. Whether it is worth the extra money pretty much depends on how many freeway miles you log, for now anyway.

I don't have it in my car but I've tried it extensively in the various loaners I've been driving. Frankly, I think AP2, to this day, is crap. Yes, it does a few things well here and there, but given that it misses basic things, such as seeing the five other vehicles stopped at the light ahead (something that AP1 does not seem to have to have an issue with), I don't think it's worth the hassle or the money. Staying alert and waiting for it to fail is a lot less satisfying than simply keeping alert by driving.

This is just my opinion and counterpoint, a lot more people seem to love it.
 
  • Like
Reactions: N5329K
Most of the time EAP on. Divided highway is very good. I have noticed 2 issues
1) if there is debris like riped up tire on the road etc. EAP will not adjust for that.
2) when navigate on autopilot is wrong and is telling you to get off the freeway and you chose not too - the car can “brake” startling the folks behind you

Other than that EAP does great (for divided highway use) - it has even learned to slow down on interchange between freeways
 
20 days with EAP and it has been pretty 'meh' for me. It works well, no glitches, but I just find it does things that I prefer to do myself. My commute is mostly divided highway with little traffic. Once in a while I do have to go other places in stop and go traffic, but very little as EV get to use express lanes here. In stop and go I can see some potential. It could be a bit smoother as it accelerates and stops too aggressively as it only looks at the car infant rather than seeing everything going on in traffic.

About the only time I use it now is when someone else is in the car and I say "look how cool this is".
 
  • Like
Reactions: N5329K and rlb4
EAP was the reason I purchased a Tesla. I don't care about an electric car vs an ICE. The Tesla is the car that is the most progressed in self-driving capabilities and they have the ability to upgrade. That was enough for me.

My wife's Mercedes has adaptive cruise control and lane keeping and I find them indispensable on both long trips and stop and go traffic. I will not buy another car that does not have those capabilities.
 
I just got the 30 day EAP trial and I've tested it out a few times and I think it's pretty amazing but just not quite sure it's right for me. I don't like how it keeps exact distance between you and the car in front of you no matter what. If you're in heavy traffic and have a bad driver in front of you who is heavy on the accelerator and brake, your car is going to end up doing the same thing. Normally in that situation I'd drive in a way where I average out the speed of the car in front and to absorb the heavy stops and starts in order to maintain more of a constant speed. Obviously I can't expect the EAP to do this, but I find it far less comfortable for that reason.

Even without EAP, I was coming from a manual transmission in my last car, so having regen braking which is basically one pedal driving on freeways is an absolute dream for stop start traffic.
 
  • Like
Reactions: Emmexx1
Normally in that situation I'd drive in a way where I average out the speed of the car in front and to absorb the heavy stops and starts in order to maintain more of a constant speed. Obviously I can't expect the EAP to do this, but I find it far less comfortable for that reason.

No, we absolutely should expect EAP to do that, it’s really not that difficult.

Both the brakes and the acceleration of the electric motor are practically flat, so allowing the car to brake and then accelerate smoothly while allowing the distance to the vehicle in front to vary within a normal/natural range should be a rather simple function of the brake pressure and motor input respectively.