Brake temps hot on downwind side.

[Escape Path]

Interesting. Reading that last paragraph I seem to have the same recollection, re: rudder input

But how does the aircraft know what "straight" is? Is it via rollout mode or via heading/track data sent to brakes?

[applecrumble]

I would think it is highly unlikely to receive any input of heading or roll out.
It will be the fact that as you touch down (with no input from the pilot) let's say there is a slight nose right movement.
The left wheel on the outside on the turn will be moving faster and therefore more brake will be applied to that wheel. (The wheel tacometer is measuring wheel speed and comparing against reference speed ADIRS speed). The right wheel is moving slightly slower as it is on the inside of the turn and so less brake will be applied thus helping steering.

[Uplinker]

But how does the aircraft know what "straight" is? Is it via rollout mode or via heading/track data sent to brakes?

Well not rollout, because this (braking) feature is available even without an ILS.

So maybe the IRS, although I cannot find whether the autobrake is unavailable if all IRSs have failed.

Or possibly wheel speeds, as applecrumble suggests, although if the runway was contaminated and the anti-skid was operating, the directional control might then go pear shaped?

[Chu Chu]

Using wheel speeds sounds a little scary. If the aircraft got off the runway heading (maybe because of momentary wheel slip, slight differences in tire wear or inflation, or just not enough autobrake authority to fully prevent weathercocking) the autobrakes would then fight to keep it on the wrong heading.

[Goldenrivett]

Thinking about it, during landing I never have to do much to keep straight in a crosswind after the autobrake cuts in, but I do have to work the rudder/NWS to keep straight during a crosswind take-off.

Hi Uplinker,

The effect you have noticed is due to more weight on the wheels because of the ground spoilers. During take off, the wing develops more lift for a given speed than the landing wing (with ground spoilers deployed) and the tyres tend to slip more in a cross wind. As they slip, you will have to "fly" more upwind with a bigger heading change.

The note about "ensuring a straight roll out" is due to equal braking effort on all wheels until one or more tyres start to slip.

So maybe the IRS, although I cannot find whether the autobrake is unavailable if all IRSs have failed.

Autobrake and antiskid need a ground speed reference to work (from IRS).

[Escape Path]

That's probably a last resort call.

It's all probably just a moot point anyway, as the brakes get hot unevenly since they are most of the time at a different wear.

Interesting discussion though

[applecrumble]

I had a look at this in the cruise yesterday and I think I was talking more about the anti skid.
Thinking about it, I think it's nothing more than just the fact that it applies even braking.
So it will be more straight than if the pilot manually brakes as it's hard for him to get exactly even pressures on both sides.