Steering Lag

[Keeney]

Interesting result of steer by wire is that it allows the driver to exceed the slew rate of the steering motor.

In a power-assist steering system, when you hit the maximum slew rate, the steering wheel gives you that feedback. You cannot turn it faster than the motors can slew the wheels because they are still connected.

This twitter post calls it lag, but I don't think that is the problem.

 

[Keeney]

Remember the old free-spinning steering wheels on early video games?

 

[Eddyrelik]

I mean cmon. Who’s going to want to turn full tilt so fast? What real world scenario would you do that? Impossible to turn that fast with regular steering rack anyway

and even playing it at 2X speed, it’s still slow motion video. So they slowed it down a ridiculous amount to show that.

 

[bertley82]

This is silly. Same old story with various FUD, just a new Tesla vehicle. Anyone that’s owned a Tesla can attest to the satisfaction. CT will be our 4th. Would never go back to Legacy.

 

[Keeney]

Impossible to turn that fast with regular steering rack anyway

One would have to assume that a modern electric power steering rack would have similar slew rate as an electric steer by wire steering rack - they appear to have have similar sized motors and mechanisms.

The problem with steer by wire is not any lack of speed on the steering rack. The problem is that the driver can always turn the steering wheel faster than the steering rack can move. With the high gain of the steer by wire at low speed, the decoupling could easily result in over-shoot in some situations.

Several early reviewers have said the steering takes a little getting used to.

 

[TickTock]

I think a better test would be to compare the speed of the turn to another vehicle (with an average strength person turning the wheel as fast as they can). Still the post is silly/FUD. Applies to a boundary condition that occurs... never. ..and even in this case it is hardly a safety issue as I am sure it travels from one extent to the other faster than I can turn my current truck (requires multiple rotations of the wheel).

 

[Deleted member 17810]

So much huge lag it had to be slowed down to 1/4 time

 

[cvalue13]

I mean, a 1/4 of a second is significant

the better (non pro) reaction times are ~0.75 seconds


A deer or a child running into the road is about 3/4 second, and then you add another 3/4 second to decide to act and move your body – that's 1.5 seconds so far. At 55 mph, the distance traveled in that period is ~120 feet. Only then does the car begin to move.

adding or subtracting a 1/4 second to those reaction times is non-trivial

I think people calling these “edge cases” fail to remember that another word for those are “emergency situations”

I’m not saying this video shows an incontrovertible problem.

only that those dismissing it out of hand aren’t really being obviously reasonable

 

[HaulingAss]

I think a better test would be to compare the speed of the turn to another vehicle (with an average strength person turning the wheel as fast as they can). Still the post is silly/FUD. Applies to a boundary condition that occurs... never. ..and even in this case it is hardly a safety issue as I am sure it travels from one extent to the other faster than I can turn my current truck (requires multiple rotations of the wheel).

Exactly!

Always dismiss, out of hand, any supposed problem of a new or different way of doing something. if that "problem" can't even be replicated by the traditional way of doing things.

It would make more sense to say a traditional steering rack has dangerous steering lag because it takes too long to go from center to full lock. And this can be a real world problem. I've driven trucks that were impossible to wrestle through tight chicanes as fast as I would have liked simply because it was impossible to put enough turns on the steering wheel that quickly.

The Cybertruck was demonstrated to go through the tight turns of the go cart track at higher speeds than a legacy truck could negotiate them, without any issues. The steering system was proven superior at fast steering transitions.

Wondering if the fact that the wheels are not always pointing exactly the direction supossed because it lags by 1/8th of second is a safety concern is really silly. Chalk it up to the finer points of the steering feel of the truck. Humans drive by experience and will quickly learn how the truck handles such quick steering inputs (if, in fact, they ever use quicker steering inputs than would be possible in other vehicles).

 

[Cold.Truth]

“Tesla Steer by wire Lag is a safety issue!” - I’m sorry, but this “Facts Chaser” post holds all the hallmarks of fear p*rn clickbait by FUD folks or worse, $TSLA short-sellers trying to materially damage the company and its investors. Note how the immediate assumption is this steering “lag” is a safety issue and needs to be fixed, similar to the “massive Tesla safety recalls” that amounted to a meaningless NSHTA mandate for a larger font size on one screen – which can be pushed out via OTA update. No mention of how SBW might enhancesafety…



“CT stuck in the snow, needs Ford F250 to pull it out!” – I challenge anyone to google “F150, F250, or RAM stuck” and see what they find. User error, wrong tires, challenging terrain/conditions, etc. Yes, the CT can get stuck. So can any truck.



“CT will chop your fingers off!” – How many vegetables are people going to stick in door and frunk seams? It’s ludicrous. Have them do it with a Ford F-150 and see what happens. Does an F-150 have the door jam blockers that prevent a recently-opened door from closing unless deliberately pulled open first? Where is the outrage and fear that the blockers are missing from ICE trucks and cars, huh?



“CT front end is hazardous to pedestrians!” - Compare front ends of a CT with an ICE truck and tell me which seems more concerning to impact. Park them nose to nose. I’ll pick the CT over the RAM all day long.



All this is being done for click$ and to damage the people who invested in $TSLA.



I am not anti-ICE or anti-Ford. I own a 1987 air-cooled stick shift ICE car and a 2016 Triumph Bonneville with no plans to sell either. However, I’ve been driving Teslas since 2013 – so I’m big EV fan with his feet in both worlds (unlike Facts Chaser, I strongly suspect).

 

[Deleted member 17810]

I mean, a 1/4 of a second is significant

the better (non pro) reaction times are ~0.75 seconds


A deer or a child running into the road is about 3/4 second, and then you add another 3/4 second to decide to act and move your body – that's 1.5 seconds so far. At 55 mph, the distance traveled in that period is ~120 feet. Only then does the car begin to move.

adding or subtracting a 1/4 second to those reaction times is non-trivial

I think people calling these “edge cases” fail to remember that another word for those are “emergency situations”

I’m not saying this video shows an incontrovertible problem.

only that those dismissing it out of hand aren’t really being obviously reasonable

It's "lag" in the back end of the motion.

Not on the front end.

 

[Derektsla]

I would wonder if it’s way more likely for someone to swerve, overcompensate back, lose control, and hurt themselves or others. Rather than “not swerve enough.”

One could speculate this slight delay could prevent some loss of control and have a net positive effect.

But who knows. That’s why speculation is stupid, especially when disguised as facts.

 

[cvalue13]

It's "lag" in the back end of the motion.

Not on the front end.

turn wheel hard right, wheel/vehicle responds and turns right shortly after

and vice versa

unclear what your comment means

 

[JBee]

I think there are three questions that needs to be asked here to identify the difference.

1) if a full lock maneuver to avoid an obstacle occurs seldomly, then
2) the more meaningful question would be how much does it lag to reach a steering angle to achieve the maneuver, and
3) does the SbW steer faster lock to lock, than a able bodied person with a mechanical setup.

If 3) is true the other two are within reason, just as safe. If 2) the steering angle required is only small, then 3) is less important and makes 1) inconsequential.

But, I have a few caveats here.

a) does the steering always respond at the same speed (no it doesn’t, it's speed dependent)
b) does the person always know how it will respond at every given speed
c) does it always respond at the same rate whilst emergancy braking
d) can ESP/ABS compensate for SbW lag by differential braking or torque vectoring steering with CB
e) what is the traction limit of the tyre and its rate of response in comparison to the steering SbW - the tyre always lags the wheel steering angle
f) given the small size of the 48V battery, and the two DC converters, is there are limit to how many time you can do lock to lock before the battery is flat and/or DC converters get warm? Does this introduce further inconsistencies and lag?

Overall, I'm pretty sure they have done their numbers, but where is the limit, and are they at least using ABS as a complete rack failure backup? They could even get FSD to steer then using the brakes to come to a safe stop, with even no steering column input or SbW at all. Note it is not only the steering angle that turns the vehicle, but in particular the difference in each wheel rpm (aka drifting).

 

[Deleted member 17810]

turn wheel hard right, wheel/vehicle responds and turns right shortly after

and vice versa

unclear what your comment means

It's logarithmic that I see in the sloooow motion video, that I see.