Firefly software snafu sends Lockheed satellite on short-lived space safari

A software error on the part of Firefly Aerospace doomed Lockheed Martin's Electronic Steerable Antenna (ESA) demonstrator to a shorter-than-expected orbital life following a botched Alpha launch. According to Firefly's mission update, the error was in the Guidance, Navigation, and Control (GNC) software algorithm, preventing …

[usbac]

"former Voyager scientist Garry Hunt questioned if the commercial spaceflight sector of today would take the same approach to quality as the boffins of the past."

No, they will not.

Too many years of "If it sort-of works, ship it. We can always send out (umpteen million) patches later". The skills to write the kind of code Garry Hunt is referring to have been long lost.

I does seem that SpaceX manages a little better than most, somehow.

[NapTime ForTruth]

Exactly this...

Earlier space efforts were built on rigor, lengthy and specific and detailed checklists that were themselves built on checklists, and everyone involved was focused on getting as close to perfection as humanly possible. It didn't hurt that governments and the populace were both literally and figuratively invested in the work and the outcomes.

We had something to prove, part of which was that we could do extraordinarily difficult things and get them right the first time much more often than not.

Space is expensively unforgiving, and frowns on the wasteful, half-assed "fail fast and iterate" model of development. It is no place for dabblers and dilettantes.

[Bebu]

Re: Exactly this...

《Earlier space efforts were built on rigor, lengthy and specific and detailed checklists that were themselves built on checklists, and everyone involved was focused on getting as close to perfection as humanly possible.》

This was brought home to me years ago when visiting the "Powerhouse Museum" in Sydney where there was some Gemini hardware on display that a visitor could inspect closely.

I was really surprised at the precision and fineness of the workmanship. There were wired seals everywhere which presumably denoted completion and provenance.

A different culture and attitude.

Another thing I noticed at the time was that much of what I could see wasn't all that complicated. So perhaps St Exupery's aphorism from Terre des Hommes (Wind, Sand and Stars):

《 Il semble que la perfection soit atteinte non quand il n'y a plus rien à ajouter, mais quand il n'y a plus rien à retrancher》

was implicitly understood by those engineers, scientists and technicians.

[fishman]

Re: Exactly this...

"Earlier space efforts were built on rigor, lengthy and specific and detailed checklists that were themselves built on checklists, and everyone involved was focused on getting as close to perfection as humanly possible. It didn't hurt that governments and the populace were both literally and figuratively invested in the work and the outcomes."

This is the type of arguement Boeing used to try to cut SpaceX out of the commercial crew program. We all know how well that worked out.

[rcxb]

I don't see why private space launch companies are getting all the flack. NASA's SLS and Artemis program are ridiculous square-peg, round-hole plans to cobble together the biggest political contributors' products into a basically unworkable layer cake of a moon program...

"at least 15 launches will be required to refuel HLS in orbit per crewed mission"

[vtcodger]

Move fast, Break things

Move fast. Break things. ... Oh sh...

[tony72]

The Voyager engineer's comment is not really fair to be honest, since he's not comparing like with like (or maybe the quote is out of context). You can afford to spend decades of time and billions of dollars trying to get a mission right on the very first try when it's a once-in-a-generation science mission to advance the boundaries of mankind's knowledge (and even then, you won't always succeed). However it makes no sense to do that when you're building a commercial launcher that's going to be launching regularly, and needs to make a profit for its builder to survive - you would go out of business before you even started. As SpaceX and Rocket Lab have shown, when you're going to be building and launching many of a rocket, an iterative approach can work, and indeed makes a lot of sense, and you can afford a few failures at the beginning while you work out the bugs.

[Headley_Grange]

Using customers as beta-testers for iPhone code is one thing, but I don't think that satellite makers are going to be happy about it.

[I ain't Spartacus]

Early rocket launches are free, or super-cheap for a good reason. You can't get payload insurance either. It's very much done at your own risk.

[Caver_Dave]

There are still vestiges of ‘proper’ software development in the certifiable software world.

All requirements, specifications, architectures, code, test procedures, test code and test results, reviewed by at least 3 different people/teams against previously reviewed checklists. Plus further independent reviews by the customers teams and the certification authority.

However, even in this world there is talk of Minimum Viable Product and Agile, with capability creep approaching the specified goals at some point in the future.

Hopefully, I will be long gone before this comes to fruition.

[fishman]

"I does seem that SpaceX manages a little better than most, somehow."

SpaceX has over 200 consecutive successful booster *landings* with the Falcon 9. One can argue that the the Falcon 9 is the safest rocket ever - the current version of the Falcon 9, Block 5, has 245 consecutive launch successes.

I'd imagine that the cost to insure a satellite is going to be quite a bit cheaper if it is being launched on a Falcon 9.

[Anonymous Coward]

ITYM...

if ( orbit = wrong) return *earth;

[Steve Button]

Re: ITYM...

If you deliberately put = instead of == then you deserve a thumbs up, but if it's a bug you deserve a thumbs down for trying to be clover.... dammit clever.

Now own up.

[trindflo]

Re: ITYM...

I take it you pronounce '*' as 'splat'

[Yet Another Anonymous coward]

So in aerospace speak

ESA's GNC of RNC led to WTF

[bemusedHorseman]

In spaceflight, the most terrifying word you can say is "whoops".

[John Brown (no body)]

"Fly the Lightning"

The launch on December 22, 2023 – dubbed "Fly the Lightning"

You just know a launch company isn't planning on any significant number of launches in the near future when they still can find the time for cutesy names for each launch :-)

[The Oncoming Scorn]

Obligatory....Browncoat Icon

Zoe: Planet's coming up a mite fast.

Wash: That's just 'cause— I'm going down too quick. Likely crash and kill us all.

Mal: Well, that happens, let me know.

[RockBurner]

Re: Obligatory....Browncoat Icon

I was going to comment that the code obviously needed a good "scrub", but wasn't sure if that might have been a somewhat overly oblique cross-reference.

[Hurn]

4x Alpha launches? When's the Beta launch?

Apparently, "Alpha" is the name of the 2 stage launch vehicle, made by Firefly. This was the 4th launch.

(Had to look that up, as was not clear, from the context.)