This is just your assumption.
So far, AP2.0 is more or less at parity with AP1.0, despite a huge difference in the hardware.
Let's not judge until we are able to test HW2.0 FSD cars against HW2.5 FSD cars under identical conditions.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
HW2.5 capabilities
- Thread starter verygreen
- Start date
-
- Tags
- Autonomous Vehicles
This is just your assumption.
So far, AP2.0 is more or less at parity with AP1.0, despite a huge difference in the hardware.
Let's not judge until we are able to test HW2.0 FSD cars against HW2.5 FSD cars under identical conditions.
Of course it is my assumption - or should I say analysis.
But the thing is: why make these changes? I can understand generational shifts, but why such a small (yet significant) change?
It certainly seems plausible these changes might be required by e.g. regulators down the road causing a significant difference in capability between AP 2.0 and 2.5.
Of course it is also possible neither will reach the goals of self-driving and a generational shfit is required (e.g. P90DL V1/V3 vs. P100D on reaching stated performance goals).
Sorry, in the absence of facts, it is an assumption.
You assume you know the feature set of the new radar they are using. You assume you know how the second Parker chip will be used. You assume the redundancy is for future regulation, when it clearly is insufficent to meet any true measure of redundancy. Etc.
AP2 is not at parity with AP1. AP2 can not lane change on any multi-lane road. It does not display vehicles on the IC in the say comprehensive way and it does not classify vehicles. It doesn't read street signs or speed limit signs. It is clearly the lesser ADAS.
I disagree with your characterization. It assumes a lot way too definitely compared to what I'm thinking or have not made up my mind on or know is uncertain.
I am considering AP2 being insufficient as one possible reason/outcome, but definitely not the only one. Hence analysis fits better, I am analyzing the tidbits we have and offering rationale on them. I am not assuming any of the reasons/outcomes as definitive. I have given my rationale on why I consider those reasons/outcomes - feel free to argue them.
You on the other hand offer nothing but, well, hope to this thread.
Let's hope it is not misplaced.Agreed, though I think everyone agrees that is irrelevant to the context here - I think nobody assumes AP2 can not do a lot better than it currently does and a lot better than AP1, eventually (if given software).
The context is a discussion on whether or not HW 2.5 provides a significant delta in favor of full self-driving regulatory approval compared to HW 2.0. I argued the changes sound such that it certainly might. Mind you, I'm not saying definitely, just saying they are the sort that might give HW 2.5 a definite edge, especially taken together with the recent news that FSD AP2 announcement was met with disbelief by Tesla's own engineers at the time...
Presumably a better radar will provide better data for TeslaVision to do a better job (especially, apparently, with pedestrian detection/recognition). Further, a biometric camera will likely eliminate the "nag" that is present in AP1/2. The rest of the information is pure speculation and therefore analysis of that is fraught with error.
"full regulatory approval" is a red herring. No regulator has withheld approval nor are there defined regulations that Tesla is aspiring to meet. Thus, from a legal perspective, nothing stops Tesla from turning on the FSD code and seeing what happens other than personal injury and product liability concerns from defective software that was only good for a couple demo videos.
Isn't this exactly what Tesla is constantly doing? It's called Agile development. Making small changes to improve the car and/or cut costs.
Maybe the reason they changed radar units is because the new one is cheaper because they got a deal as part of the Model 3 volume and has nothing to do with its capabilities.
Or maybe Tesla has no idea what it is doing and its just throwing punches in the dark trying to score a KO. Assuming a better functioning unit is cheaper is probably the worst assumption one could make. Continental isn't stupid nor is Tesla really producing the Model 3 in volume. Its a joke to assume they are getting bulk discounts as compared to actual OEM volume manufacturers that put out 10M units a year (while Tesla still fails to produce >100k).
Discounts are done based on order volume over time, not what your current delivery/usage is. And I wasn't comparing what other OEMs are paying. I'm saying that for the new radar unit they probably have a contract for ~600k units/year, where with the old Bosch unit they were only ordering ~100k units/year. So yes, I suspect that with 6 times the volume that they will pay less now than they were before.
Its clear that you do not know how a business operates. Better tech is never cheaper. Just let the disagree fly, it won't change reality.
It is a new thing for their Autopilot platforms that have remained statics hardware targets thus far within their generations.
It is natural to ponder what triggered it, especially given that it looks more a hotfix to AP2 than a generational shift. And there are the news emenating from within Tesla of disbelief in FSD claims.
This is not just a new radar, but as host of cabling and power/fuse and possibly performance differences.
I disagree with your characterization. It assumes a lot way too definitely compared to what I'm thinking or have not made up my mind on or know is uncertain.
I am considering AP2 being insufficient as one possible reason/outcome, but definitely not the only one. Hence analysis fits better, I am analyzing the tidbits we have and offering rationale on them. I am not assuming any of the reasons/outcomes as definitive. I have given my rationale on why I consider those reasons/outcomes - feel free to argue them.
You on the other hand offer nothing but, well, hope to this thread.
Whereas you spread your pessimistic opinion far and wide, trying to legitimise assumption as "analysis"
Well, I'm sure the answer is somewhere in the middle.Back on topic. I am curious as to the "primary" and "secondary" wiring to the new radar. Given Elon's many comments about wanting to simplify wiring looms (speaking about M3, but principals apply equally to MS and MX) it is interesting that they are adding to the loom without any obvious need or benefit.
I haven't been able to find any current Conti ARS datasheets online, but instrustrial versions come with a single CAN bus. Are the new primary/secondary cables for CAN or something else?
I haven't been able to find any current Conti ARS datasheets online, but instrustrial versions come with a single CAN bus. Are the new primary/secondary cables for CAN or something else?
Whereas you spread your pessimistic opinion far and wide, trying to legitimise assumption as "analysis"
Well, you sure are doing your best to delegitimize my analysis...
To be clear: other alternatives than the pessimistic one are included in my view of how things may unfold. I tried to convey that as well.
lunitiks
Cool James & Black Teacher
It's not like there's two wiring harnesses from the new radar. Still just one connector. But there are more individual wires to this cable, namely a doubling of CAN-signal wire. They both go directly to AP2.5ECU, so no more info than the wording "Prim" and "Sec". Exciting tho
There's a redundancy requirement for L5. To me that explains the extra wiring and any other electronic backups.
If someone wants to chase down the dates of the NHTSA's L5 guidelines vs. AP 2.0 release, then we may find that AP 2.0 preceded the new NHTSA guidelines. Not sure, but that would explain a lot of these moves.
My take is that AP 2.5 is about ride-sharing and meeting this nebulous redundancy "requirement". I wrote a member article to that effect a couple of weeks ago.
Given the new information about an upgraded radar do we know whether or not the old radar supported double CAN-bus signaling? If it didn't, then this move could again be explained under the "redundancy" clause, and it's upgraded specs might have just been a bonus.
If someone wants to chase down the dates of the NHTSA's L5 guidelines vs. AP 2.0 release, then we may find that AP 2.0 preceded the new NHTSA guidelines. Not sure, but that would explain a lot of these moves.
My take is that AP 2.5 is about ride-sharing and meeting this nebulous redundancy "requirement". I wrote a member article to that effect a couple of weeks ago.
Given the new information about an upgraded radar do we know whether or not the old radar supported double CAN-bus signaling? If it didn't, then this move could again be explained under the "redundancy" clause, and it's upgraded specs might have just been a bonus.
lunitiks
Cool James & Black Teacher
Mjeh... Tesla actually uses the word "Redundant" in the wiring diagrams like for example Gen 3. Power Steering Redundant power.
But the new radar has "Primary" and "Secondary" CAN-bus wires. This doesn't have to mean the secondary is for redundancy. Like the aforementioned paper indicates:
Similar threads
- Article
