50% of A320s need a software roll back
Discussion
https://ad.easa.europa.eu/ad/2025-0268-E
Airworthiness Directive has dropped, must be done before next flight although you can fly home empty of passengers if it's stuck somewhere awkward.
Airworthiness Directive has dropped, must be done before next flight although you can fly home empty of passengers if it's stuck somewhere awkward.
Blib said:
DeejRC said:
It s interesting.
Amusingly ironic - to me anyway - for this to happen on the day I had to step back into being FSW. Fortunately this ain t one of my gigs so not my problem
FSW......?Amusingly ironic - to me anyway - for this to happen on the day I had to step back into being FSW. Fortunately this ain t one of my gigs so not my problem
FSW........?
Flying Somewhere Wrong !!!!!!!
"What does FSW mean?
(AI Overview)
FSW is a versatile acronym with several meanings depending on the context, including Friction Stir Welding, a manufacturing process; Federal Skilled Worker Program, for Canadian immigration; and Family Support Worker, a social services role. Other meanings include Female Sex Worker, forward swept wing, and the unit feet sea water used in diving."
Interesting.
gotoPzero said:
They have yet to calculate the time needed but around 6000 airframes require a software roll back - I cant see it being a 2 minute job.
Thats half of the global fleet apparently. Hopefully wont cause significant delays.
Doesn't sound terribly onerous, notwithstanding the number of aircraft it needs doing to; the directive implies it's a 'remove control module and replace with serviceable module with previous software iteration'. Without knowing how many surplus modules are in the maintenance system and how quick it is to wind the software back it's impossible to tell whether it's going to be a sThats half of the global fleet apparently. Hopefully wont cause significant delays.
tshow or a momentary whoopsie that will be forgotten next week. It's a bit strange. All the news reports imply that a "intense solar radiation is corrupting data", yet it's fixed by rolling back to a previous software version, rather than fitting massive amounts of shielding around something?
Two possibilities I can think of:
The new software had a bug where it's failing to check CRCs or similar on incoming messages
Or, somewhere there's a sensor that measures this radiation, and there's a bug when the values go outside an expected range
Other suggestions?
Two possibilities I can think of:
The new software had a bug where it's failing to check CRCs or similar on incoming messages
Or, somewhere there's a sensor that measures this radiation, and there's a bug when the values go outside an expected range
Other suggestions?
Bradgate said:
And someone has pressed +300% on the photocopier.
It’s the solution! Bin off all of tomorrow’s Heathrow to Glasgow flights, put the punters in one A380 and problem solved!Though the reality is probably more like the poor sods travelling from Manchester, Leeds, Newcastle etc to Heathrow will be going on a coach or just getting cancelled and rebooked for an unknown date. Flightradar isn’t showing any signs of mass cancellations at the moment, will be interesting to check back around 6am tomorrow
silentbrown said:
It's a bit strange. All the news reports imply that a "intense solar radiation is corrupting data", yet it's fixed by rolling back to a previous software version, rather than fitting massive amounts of shielding around something?
Two possibilities I can think of:
The new software had a bug where it's failing to check CRCs or similar on incoming messages
Or, somewhere there's a sensor that measures this radiation, and there's a bug when the values go outside an expected range
Other suggestions?
Solar radiation issue to me suggests someone disabled CRC checks on some memory as that's just the sort of situation where you'd get bit flipping and failure. Message checks would be a downstream failure. But could be transceiver or all sorts of other stuff glitching and not being validated.Two possibilities I can think of:
The new software had a bug where it's failing to check CRCs or similar on incoming messages
Or, somewhere there's a sensor that measures this radiation, and there's a bug when the values go outside an expected range
Other suggestions?
Not something I'd have expected to change in an update though, the hardware feature setup is not the sort of thing that gets fiddled with once it's working as its hard work - ever tried to reverse engineer a manufacturer's BSP blob to work out how to get the right bits into a certified product?
Having applied some serious abuse to avionics software in a previous life, you can get all sorts of stuff to happen if it's outside the environment it was assumed to exist in. You could make it do lots of weird things if data timing was off or the values weren't behaving as the design expected. So if something got fiddled and 'impossible' things started getting through you could have a fun ride.
JoshSm said:
Solar radiation issue to me suggests someone disabled CRC checks on some memory as that's just the sort of situation where you'd get bit flipping and failure. Message checks would be a downstream failure. But could be transceiver or all sorts of other stuff glitching and not being validated.
Not something I'd have expected to change in an update though, the hardware feature setup is not the sort of thing that gets fiddled with once it's working as its hard work - ever tried to reverse engineer a manufacturer's BSP blob to work out how to get the right bits into a certified product?
Having applied some serious abuse to avionics software in a previous life, you can get all sorts of stuff to happen if it's outside the environment it was assumed to exist in. You could make it do lots of weird things if data timing was off or the values weren't behaving as the design expected. So if something got fiddled and 'impossible' things started getting through you could have a fun ride.
Or, they've cocked up the triple-redundant voting bit somehow>Not something I'd have expected to change in an update though, the hardware feature setup is not the sort of thing that gets fiddled with once it's working as its hard work - ever tried to reverse engineer a manufacturer's BSP blob to work out how to get the right bits into a certified product?
Having applied some serious abuse to avionics software in a previous life, you can get all sorts of stuff to happen if it's outside the environment it was assumed to exist in. You could make it do lots of weird things if data timing was off or the values weren't behaving as the design expected. So if something got fiddled and 'impossible' things started getting through you could have a fun ride.
Gassing Station | Boats, Planes & Trains | Top of Page | What's New | My Stuff