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

Tesla new steer-by-wire, that'll be a HELL NO for me....

This site may earn commission on affiliate links.
from all reports of the Cyber Truck - the steering is a game changer. There are some descriptions of the physical build of the install which looks to be pretty robust. I've had a tie rod end snap on a car, so even mechanical systems fail. I'd like to try it!
 
  • Like
Reactions: ArchHamster
from all reports of the Cyber Truck - the steering is a game changer. There are some descriptions of the physical build of the install which looks to be pretty robust. I've had a tie rod end snap on a car, so even mechanical systems fail. I'd like to try it!
A steer by wire system would still be prone to tie rod failure.

I think a software solution could be okay for steering. But I don't trust Tesla's software engineers for mission critical systems. Just about everything they've released so far has been half baked from a software standpoint
 
A steer by wire system would still be prone to tie rod failure.

I think a software solution could be okay for steering. But I don't trust Tesla's software engineers for mission critical systems. Just about everything they've released so far has been half baked from a software standpoint
You do realize the accelerator is pure "speed by wire" right?
 
It’s far simpler than steering, and losing acceleration is not as critical as losing steering.
Really? So suddenly going to full throttle isnt dangerous?

-- Accelerator-by-wire: Sensor detects pedal movement and sends information to the computer. The computer sends commands to the motor control units to alter the power to the motor(s). The motors act via mechanical linkages to change the speed of the car.

-- Steer-by-wise: Sensor detects steering wheel movement and sends information to the computer. The computer sends commands to the steering control unit to move the steering actuators. The actuators act via mechanical linkages to change the direction of the car.

Not sure how these are so different to warrant your distrust of one and trust of the other.
 
  • Helpful
Reactions: Krugerrand
Really? So suddenly going to full throttle isnt dangerous?

-- Accelerator-by-wire: Sensor detects pedal movement and sends information to the computer. The computer sends commands to the motor control units to alter the power to the motor(s). The motors act via mechanical linkages to change the speed of the car.

-- Steer-by-wise: Sensor detects steering wheel movement and sends information to the computer. The computer sends commands to the steering control unit to move the steering actuators. The actuators act via mechanical linkages to change the direction of the car.

Not sure how these are so different to warrant your distrust of one and trust of the other.
You can always overpower acceleration with your brakes and steer to avoid obstacles worst come to worst. And the typical failure mode would be loss of power rather than stuck full throttle.

If your steering goes in the middle of a drive, you’re potentially ****ed. It’s not remotely comparable.

Throttle by wire was also used in the industry decades before Tesla existed and was already mature tech. Steer by wire is brand new and Tesla’s own implementation. Time and time again Tesla has shown they’re willing to release immature and buggy products into the wild. So yeah, I think my lack trust is warranted.
 
You can always overpower acceleration with your brakes and steer to avoid obstacles worst come to worst. And the typical failure mode would be loss of power rather than stuck full throttle.

If your steering goes in the middle of a drive, you’re potentially ****ed. It’s not remotely comparable.

Throttle by wire was also used in the industry decades before Tesla existed and was already mature tech. Steer by wire is brand new and Tesla’s own implementation. Time and time again Tesla has shown they’re willing to release immature and buggy products into the wild. So yeah, I think my lack trust is warranted.
Hmm have you ever tried overpowering a full-throttle car with brakes? If steering fails mid-drive what are you going to do? Brake like crazy. The same you would do if the car suddenly zooms forward. Which failure mode results in a faster stop? Steering failure of course, since the car is not fighting the braking. So your logic is flawed, its throttle control failure that is the most dangerous.

And you entirely miss my point. The stack for throttle-by-wire and steer-by-wire is essentially identical in an EV. So why do you trust Tesla with throttle-by-wire and not steering-by-wire? Comparing to ICE throttle-by-wire is invalid since that is a totally different stack. If you dont trust Tesla with steering-by-wire you should NOT trust them with throttle-by-wire either, by your own logic.
 
Hmm have you ever tried overpowering a full-throttle car with brakes? If steering fails mid-drive what are you going to do? Brake like crazy. The same you would do if the car suddenly zooms forward. Which failure mode results in a faster stop? Steering failure of course, since the car is not fighting the braking. So your logic is flawed, its throttle control failure that is the most dangerous.

And you entirely miss my point. The stack for throttle-by-wire and steer-by-wire is essentially identical in an EV. So why do you trust Tesla with throttle-by-wire and not steering-by-wire? Comparing to ICE throttle-by-wire is invalid since that is a totally different stack. If you dont trust Tesla with steering-by-wire you should NOT trust them with throttle-by-wire either, by your own logic.
If you think the “stack” is at all similar, you really have no idea how these things work. They’re fundamentally different, and throttle is drastically simpler than a steer by wire system.

And like I said, in the event of throttle failure, you can brake *and* steer. But losing the ability to control the direction of the car is far more catastrophic and much more likely to lead to an accident. Additionally as I stated earlier but you conveniently ignored to suit your narrative, the failure mode for throttle failure is far more likely to be loss of power rather than being stuck in full throttle. Much safer from that standpoint too than what you’re implying.
 
Based on reviews I've seen so far, the tech is amazing to me, I for sure want it, and hope it becomes available for the X in the near future.
I'm not concerned about the failure of the system, I have confidence that Tesla's engineering team have done their due diligence with back ups incase of failures. I'm more concerned now that the car can go rogue and fully be taken over. There's nothing mechanical now for one to stop it. Other than opening the emergency door and jumping out. But that's unlikely right? :confused:
 
Additionally as I stated earlier but you conveniently ignored to suit your narrative, the failure mode for throttle failure is far more likely to be loss of power rather than being stuck in full throttle. Much safer from that standpoint too than what you’re implying.
I didnt ignore it, since you presented no rationale for this claim. Most of your arguments to date have been about the issues around the software component and how that adds a latyer of uncertainty around the "by wire" concept. Which is why I addressed the similarity between both the accelerator and steering stacks when it comes to this. And, if you ARE indeed concerned about the software stack, I cannot see any way you can assume that a failure of the software component is "more likely" to result in a loss of power rather than sudden acceleration. To be sure, a rigorous software component SHOULD be designed to fail in a safe way, but your whole argument is that you do NOT trust the software to do this. So on the one hand you claim a software stack IS safe and on the other that a similar stack is NOT safe, which doesnt seem to me very logical.

And if not the software, then what? Your point is not to trust by-wire steering, which differs primarily form mechanical linkages by the introduction of a software component and related actuators. In actuality, I think you ARE right to be cautious, my point is not to argue that steer by wire is safe, but that you cannot on the one hand say its unsafe while you happily drive a car using throttle-by-wire. And your argument about throttle-by-wire being tested for years is of course invalid since we are dealing with EV and not ICE engines, which differ fundamentally in the throttle stacks (and in fact have a much more complex software stack than steer-by-wire).
 
I didnt ignore it, since you presented no rationale for this claim. Most of your arguments to date have been about the issues around the software component and how that adds a latyer of uncertainty around the "by wire" concept. Which is why I addressed the similarity between both the accelerator and steering stacks when it comes to this. And, if you ARE indeed concerned about the software stack, I cannot see any way you can assume that a failure of the software component is "more likely" to result in a loss of power rather than sudden acceleration. To be sure, a rigorous software component SHOULD be designed to fail in a safe way, but your whole argument is that you do NOT trust the software to do this. So on the one hand you claim a software stack IS safe and on the other that a similar stack is NOT safe, which doesnt seem to me very logical.

And if not the software, then what? Your point is not to trust by-wire steering, which differs primarily form mechanical linkages by the introduction of a software component and related actuators. In actuality, I think you ARE right to be cautious, my point is not to argue that steer by wire is safe, but that you cannot on the one hand say its unsafe while you happily drive a car using throttle-by-wire. And your argument about throttle-by-wire being tested for years is of course invalid since we are dealing with EV and not ICE engines, which differ fundamentally in the throttle stacks (and in fact have a much more complex software stack than steer-by-wire).
I mean it’s pretty obvious that the most likely failure would be that the thing just won’t work (software crash or whatever causing a failure to start and power not to be delivered). In the case of throttle that means no power. In the case of steering, that means no steering. Doesn’t take a genius to figure out which one is more dangerous.

You also keep repeating as nauseum that the throttle is more complicated and the decades of validation pre EV is somehow not applicable. That’s just nonsense and tells me you have zero clue as to how any of these subsystems work.
 
Based on reviews I've seen so far, the tech is amazing to me, I for sure want it, and hope it becomes available for the X in the near future.
I'm not concerned about the failure of the system, I have confidence that Tesla's engineering team have done their due diligence with back ups incase of failures. I'm more concerned now that the car can go rogue and fully be taken over. There's nothing mechanical now for one to stop it. Other than opening the emergency door and jumping out. But that's unlikely right? :confused:
Teslas shitty software engineering is much more likely to be a problem than a skynet scenario..
 
NASA has been operating a fly-by-wire F-8 since 1972, and Airbus has been fly-by-wire since 1988. Compared to aircraft, steer-by-wire is a cake walk. Now of course how it is implemented is key, but there is plenty of history to show it can be done safely.
 
If you think the “stack” is at all similar, you really have no idea how these things work. They’re fundamentally different, and throttle is drastically simpler than a steer by wire system.
Let's roughly sketch out what each stack has to do:

Sterr-By-Wire:
-- Read steering wheel position as input
-- Observe current vehicle speed (wheel rotation) as input
-- Read current steering 'feel" option chosen by driver
-- Compute desired rack position
-- Drive rack actuator to desired position, with appropriate ease-in/ease-out, taking into account vehicle speed as needed
-- Monitor rack position (preferably using dual independent sensors for redundancy and safety)

Throttle-By-Wire:
-- Read accelerator pedal position
-- Read brake pedal position
-- Read current throttle options, including mode (chill, sport etc), regen setting, creep/roll mode etc
-- Observe current vehicle speed (wheel rotation) as input
-- Compute desired acceleration profile based on all these settings
-- Control inverters on motors to map desired profile to torque, taking into account speed and driver options
-- Handle blending front/rear motor torque
-- Handle torque vectoring based on wheel spin
-- Apply brakes as necessary for additional torque vectoring for vehicle stability
-- Handle seamless transition from acceleration to regen (flipping inverter modes without creating jolts to car/occupants)
-- Handle creep simulation if enabled
-- Handle coasting simulation if enabled (like ICE automatics)
-- Apply brakes to simulate regen mode when necessary (battery 100% or too cold)
-- Apply brakes and blend out motor torque as car comes to a halt, release brakes when car is told to move
-- ...

By my reckoning, the throttle stack probably has 10x more lines of code than the steer stack. In fact, I'm pretty impressed by how smoothly the stack handles all the complexities of blending torque/regen and brakes into a near seamless throttle experience. There is a LOT of complexity going on that none or use are aware of (or need to be), but that doesnt mean it doesnt exist.
 
Based on reviews I've seen so far, the tech is amazing to me, I for sure want it, and hope it becomes available for the X in the near future.
I'm not concerned about the failure of the system, I have confidence that Tesla's engineering team have done their due diligence with back ups incase of failures. I'm more concerned now that the car can go rogue and fully be taken over. There's nothing mechanical now for one to stop it. Other than opening the emergency door and jumping out. But that's unlikely right? :confused:

There is always the P-button ;) for as long as they don't remove all the controls ;)
 
I mean it’s pretty obvious that the most likely failure would be that the thing just won’t work (software crash or whatever causing a failure to start and power not to be delivered). In the case of throttle that means no power. In the case of steering, that means no steering. Doesn’t take a genius to figure out which one is more dangerous.

You also keep repeating as nauseum that the throttle is more complicated and the decades of validation pre EV is somehow not applicable. That’s just nonsense and tells me you have zero clue as to how any of these subsystems work.
"software crashes or whatever" and it magically stops accelerating? No, unless you build in safety systems the hardware will simply continue at the same speed (or do you think the software has to keep "nudging" the car forward). Sure, some bugs cause a software crash, but many (most) do not, and these bugs could cause the car to do any crazy thing you can think of, INCLUDING shooting forward at full acceleration.

As for stack compelxity, go read my other post. And yes, the ICE throttle-by-wire similarity ends when the throttle input sensor gets its data to the software stack. EVERYTHING after that (small) functional overlap is completely different in an EV.

Since this thread now seems to be deteriorating into ad hominem silliness, I will rest here. I've made my point.
 
I mean it’s pretty obvious that the most likely failure would be that the thing just won’t work (software crash or whatever causing a failure to start and power not to be delivered). In the case of throttle that means no power. In the case of steering, that means no steering. Doesn’t take a genius to figure out which one is more dangerous.

You also keep repeating as nauseum that the throttle is more complicated and the decades of validation pre EV is somehow not applicable. That’s just nonsense and tells me you have zero clue as to how any of these subsystems work.
How did you ever separate from your typewriter and migrate to a laptop? Your transition from the Blackberry to iPhone must have been hell (how could one ever type without the physical keyboard?). I’m with you, change is hard. Progress is hell. Thomas Edison should have never invented the lightbulb (what was ever wrong with candles?).
 
  • Funny
Reactions: drtimhill
How did you ever separate from your typewriter and migrate to a laptop? Your transition from the Blackberry to iPhone must have been hell (how could one ever type without the physical keyboard?). I’m with you, change is hard. Progress is hell. Thomas Edison should have never invented the lightbulb (what was ever wrong with candles?).
I was not aware migrating from a typewriter to a laptop required trusting a safety critical system to a company that has a reputation for releasing half baked software to the public.
 
  • Like
Reactions: ArchHamster