Probes Point to Northrop Grumman Errors in January SpaceX Spy-Satellite Failure | Hacker News Reader

Probes Point to Northrop Grumman Errors in January SpaceX Spy-Satellite Failure

wsj.com

135 pointsuptown87 comments

There are at least two large problems relating to this inside Northrop Grumman as I see it.

One, the apparent technical mistakes that led to the failure, assuming these separate probes are correct.

Two, the apparent ethical failure and cowardice of top Northrop Grumman management, which has as far as I can tell has sat by silently letting SpaceX be blamed.

I'd be interested to know who the unnamed sources that fed the false narrative to Bloomberg and the WSJ about it being a SpaceX problem were. (I think those were the originating news organizations, but I may be remembering that wrong.)

This was a month before the Falcon Heavy launch which had a reasonable chance of a spectacular failure. Priming the public to think of SpaceX as a satellite destroyer and then a high profile failure would have been fantastic for United Launch Alliance.

I think it was pretty clear that SpaceX's customers didn't believe it as a SpaceX problem. Subsequent launches were not delayed pending investigation as they were for legitimate failures.

Or maybe it was just sloppy reporting from some guy shooting off his mouth.

The Wall Street Journal’s rocket reporter, Andy Pasztor [1], has had an anti-commercial space bias since as far back as I can remember. His sources appear to be Boeing and Lockheed Martin company men. His reporting echoes their world views, inaccuracies and all.

[1] https://www.wsj.com/news/author/1472

Looking through the recent Wall Street Journal A. Pasztor headlines that you've linked to, it seems Northrop Grumman has also been having production issues with the James Webb Space Telescope satellite. Upon further search, I haven't yet seen any articles claiming that the issues are technical mistakes, only that they were "avoidable".

Not a great PR month for NG.

The most recent GAO report (for which Northrop resisted requests to interview project staff[0]), blamed Northrop in part for the delays, citing “a manufacturing error and manufacturing process mistakes.”[1]

[0] http://spacenews.com/northrop-balks-at-agencys-request-to-in...

[1] http://spacenews.com/northrop-balks-at-agencys-request-to-in...

Plus take the fact that Lockheed/Boeing/ULA are public and SpaceX is not. Take the fall long and public enough that by the time the market would adjust their pricing it's largely been forgotten.

Yep, even before I saw your comment I guessed that the initial SpaceX hit-piece was written by Andy Pasztor.

This is how all journalism works, but Gell-Mann amnesia keeps us from realizing.

I don't think it's fair to call it an ethical failure. The mission was classified; presumably they asked their lawyers what they could say, and their lawyers told them not to say anything.

Then they went into a briefing and told reporters they needed to ask SpaceX what happened. That's not cool.

It's entirely possible they didn't officially have the telemetry to answer any questions until speaking with SpaceX. They also have a responsibility to their own investors and business to get the answers right the first time.

Then they shouldn't have blamed SpaceX.

It's unknown who blamed SpaceX; it could have been the Senator from Alabama's staff.

I think the HN crowd already knows the mission was classified. But as we saw certain aspects of the launch were allowed to be mentioned. For example, SpaceX made a statement afterward about what they knew. This aspect was not classified.

Lawyers don't just tell you what you can and cannot say. They also sometimes tell you what you should and should not say for PR purposes or to manipulate a situation.

There's plenty of room here to call this an ethical failure.

It's certainly possible, but I'm not sure I'd "presume" it. NG certainly had a motivation to cover up a failure they were responsible for, whether or not they were legally required to stay quiet.

I agree with this.

I really wonder what Zuma was capable of?

It was designed to have a very large unfoldable directional antenna to be able to transmit and receive signals in the 500Mhz - 5Ghz range, despite atmospheric attenuation. I would guess the goal was to be able to passively listen to wifi and LTE signals in hostile countries without hardware on the ground. The antenna was sensitive to shocks when folded, hence the modification of the payload adapter to avoid them.

That is awfully specific for a pseudonymous uncited post about a spy satellite. I don't think people should believe you.

Agreed. In future when posting detailed technical information about classified military operations, please do so from a verified account and provide sources and reference material. ;)

I’d say its specific because its a sensible extrapolation based on existing declassified information combined with the small amount of leaked information that has trickled out over the years.

While reading about the possible conspiracy that the failure did not happen and it was to cover the launch of a stealth satellite, this was basically the satellite payload I pictured. Based on its role as a high orbit SIGINT satellite it would have a huge antenna, huge antennas are fragile buggers to launch, thus the fancy custom payload adapter.

What?? You mean I can't pick up my wifi from the end of the garden, but there's a chance the US government is listening in from space?

Is that seriously possible?

While I can’t comment on if it would be possible from orbit, there are two components to receiving a radio signal - transmitter strength and antenna gain. These work together to provide a signal clear enough to be amplified in to something usable.

In a modern portable device the antenna is quite small and often close to other electronics, so the gain is not great. You can buy dedicated hardware that can receive over much greater distances.

Additionally device antennas are omnidirectional, so they pick up noise from every direction but the signal from only one. A directional antenna has much greater gain as it excludes most of the noise.

But isn't both WiFi and LTE encrypted?

Yes, but you can obtain other info without the keys, such as SSID locations.

Personally I’m doubtful, unless the satellite was in geostationary orbit then it wouldn’t have that long over each location. I guess if it was in a Molniya-type orbit, but that would take it quite high. Guess we’ll never know.

Given how frequently we discover major flaws in the implementation of TLS, despite how well that is scrutinised.

It's not unreasonable to believe there would similarly be major flaws in the implementation of 3G/LTE security that could be exploited.

It depends. Both WiFi and LTE can be attacked via evil-twin attacks where a base station pretends to be the legitimate one.

https://en.wikipedia.org/wiki/Evil_twin_(wireless_networks)

and:

https://www.documentcloud.org/documents/4392401-4G-LTE-attac...

Not an expert. But reasonably knowledgeable about what you could do with antenna tech and while this is possible in theory if there is only one transmitter in all of the Earth-side footprint of where that antenna points I highly doubt you could recover all the signals of all the individual WiFi transmitters in that footprint which is what would be necessary to pull a single data stream out of the mess.

Even so, if you're in a rural area and your next door neighbour is a mile away or so with the right antenna this might very well be possible. Keep in mind that antenna gain is the rough equivalent of increased transmitter power, so a very good antenna can offset a lack of power at the transmitter site.

Big antennas can pick up really small signals.

Deep space network video where they talk about the fact that Voyager I/II is NOT the toughest signal to grab (apparently Mars Exploration Rover is due to small antennas): https://youtu.be/1rCrfQUcXDI?t=188

That being said, I doubt you could pick up WiFi and dig it out of the noise at space distances.

However, you probably can relay signal intelligence from a drone or a human who is using a specific modulation known to the satellite.

Not picking up wifi in garden... could be akin to "I can't see the ground clearing while in a 747... so what can they really see from orbit?"

https://motherboard.vice.com/en_us/article/8qx54b/googles-sa...

Considering you can "legally" see "manholes and mailboxes" from orbit?

What could we see if not "limited" by laws?

Whats visible now considering that article was 2014...

Grab your tin-foil hats.

Yes. Space isn't that far away and with a big enough antenna...

Note that satellite phones and so on (tv) work like this in reverse.

USG has been listening to EM signals from space (elint) since the space age began.

Yep, https://www.cnet.com/news/new-wi-fi-distance-record-382-kilo... . If you have big enough antena and state of the art amplifiers, 500km is not unbelievable and that is high enough to orbit a satellite (for example ISS is at about 405km now). Plus, you have only about 100km of atmosphere (I might be wrong here).

It's not that simple, those point-to-point links are done with both highly directional antennae at both ends. You'd need to deal with Doppler (the satellite is moving pretty quickly, and likely in a relatively complex path relative to the transmitter), attenuation because of rainfall and other atmospheric effects, polarization, non-directional transmitter antenna and the effect of likely having 1000's of transmitters of roughly the same kind in the footprint of the satellite antenna.

It isn't impossible, but 'Yep' is a bit of a shortcut, and if they pull this off it would be quite the technological feat.

It also would mean that if this is possible that you wouldn't need anything special to implement SpaceX's vision of ubiquitous internet using LEO satellites, the antenna tech for that is quite exotic and requires a roof mounted structure, something that almost no WiFi transmitter has.

So those WiFi distance records are not at all on the same level as picking up a single WiFi signal from space, the one requires cooperation on both sides, the other has to do everything on one end (but obviously with a larger budget and some really clever people behind it). If you've ever done such a long haul WiFi link then you'll know how hard it is to line up the dishes and how easy it is to lose the connection due to wind, vibration or water in the atmosphere.

They've got billions of dollars to dump into a ginormous phased array with nearly perfect acoustic isolation, no gravity or kids tripping over wires to mess up an absurdly long baseline, custom software stack, custom steering loop, and probably even custom silicon.

I'd expect them to do substantially better than some dudes in the desert with glorified cookware. Especially when it comes to the steering loop, because if there's one thing phased arrays are really, really good at, that's it.

Why would there be nearly perfect acoustic isolation? Spacecraft have jitter from spinning reaction wheels or CMGs.

Money does not trump physical law.

Source?

This is assuming it actually failed.

https://allthingsnuclear.org/lgrego/lost-in-space-the-zuma-s...

The only reason this is even a thing is because Musk's fans are involved. Every other launch on all of wikipedia is a "Success", "Partial Failure" or "Failure" - regardless of who is at fault. Only SpaceX would get a special "Launch success", because it's always everyone's fault except ~~Tesla~~ SpaceX.

Look at Arianespace's last "Partial launch failure" -- upper stage was developed by a completely different company but it was still a launch failure.

I think Northrop just didn't want to get involved in the silliness.

Yeah, about that Atlas V failure that's recorded as a success? I guess Tesla/SpaceX is at fault for that, too.

Those 2 Soyuz launches that I edited on Wikipedia from success to failures, that must be Tesla/SpaceX's fault, too.

This is a classified satellite. There is also the very real possibility that it successfully launched, and both SpaceX and NG were paid to report that the satellite didn't make it to orbit.

That's not possible. Anyone with a decent radar or telescope can track the satellite. There's nowhere to hide in orbit. We don't know what was in the satellite but we do know it isn't there any more.

Unless that was the mission.... to try and launch an un-trackable satellite.

There's no possible way to build an un-trackable satellite that's large enough to make useful Earth observations. The basic limits of physics still apply. There's a difference between rocket science and magic.

Based on earlier discussions, I thought it was really hard to see a satellite if you don't know the orbit.

I did some calculations, and a 15m object (largest that fits in falcon 9) at 300km (lowest orbit) has an angular diameter of 0.006 degrees. That amounts to taking up about 0.000005% of the sky above the horizon.

I figured the required resolution to cover the sky well enough to see that would be ridiculous, but 20 megapixels spread over the entire sky would then be enough that this satellite could occlude an entire pixel.

Considering how many hobbyists there are, this does indeed seem feasible.

You don't have to scan the entire sky. Based on knowing the the approximate launch trajectory, there are only a limited range of orbits where the satellite could possibly be. Satellites have very limited fuel and can only make small orbital changes.

The angular diameter of stars that are light-years away are really really small. Yet they are visible to the naked eye and cameras all the time. So it's not just about the angular diameter.

But they shine a lot of light in every direction, including directly at you, the observer. So even if they are small they compensate for that by their power output and the fact that their radiation is a near perfect sphere. A satellite trying hard to hide isn't going to be nearly as friendly as that for an observer.

How are you going to power said satellite? Those massive solar arrays are going to reflect emag across the spectrum (radar returns and solar returns).

Honest question: If we can make stealth planes, why not stealth sats?

Stealth planes aren't invisible. They are still visible to radar, infrared and mark-1 eyeball.

The difference is that they attempt to make themselves look very small in the types of radar used to aim missiles. If the error bars on the location is large enough, you can't fire a missile and get a reliable kill.

The term I keep hearing is the "kill chain", in that stealth is attempting to break one of the parts in the chain.

https://en.wikipedia.org/wiki/Kill_chain

I thought that stealth planes had the radar cross-section of something like a marble. That was the first gen planes too, I think it might even be smaller now. I would have thought that's effectively close to invisible.

The problem is, radar returns decline with the 4th power of distance - so if you can detect a 1m square at 100km, you can detect a 1cm square at 10km and a 0.1mm square at 1km. Yeah, a marble is fantastic in comparison to a barn door - but that 4th power cuts both ways.

In order to observe Earth in a useful way the satellite needs large optics and/or large antennas. Changing the shape of those things or adding coatings to reduce radar reflectivity would make them ineffective as sensors.

Satellites also need power. Solar panels are big and have to point at the sun so no way to hide those. RTGs are also an option, but even with those there's going to be significant waste heat which shows up on IR sensors against the cold background.

And any large satellite will show up clearly on a regular optical telescope. Take a series of pictures over time and you can spot anything moving, either through direct imaging or because it occludes some other celestial object.

You can try via low albedo in any required wavelength, but satellites still occult stars. Reorientation for optimal cross section is the obvious way, no idea how much that buys.

> but satellites still occult stars

True, but it would be very difficult to 1) observe this unless you were looking for it directly and 2) determine the object's orbit. Since stars are point light sources, you can't even tell how fast it is moving since the size of the object could mislead your attempt to deduce this by timing the event.

You lost me on the timing. We are looking for small black rocks way past geostationary, and it's 100% at or inside that. Being a point source is what I want (since they all are). If it's in orbit, the speed bounds are already known, it's somewhere between 0 and as slow as practical, the lower the faster; faster is worse for an earth pointing platform. Once you know something is there, isn't that all you really need? Why does the speed matter for detection? An adversary can (presumably) almost immediately target that location if they are willing to admit they saw it. I don't feel like I'm saying anything non-obvious. It's way easier to hide in the atmosphere than outside of it.

If Zuma made it, and it's not found by the people looking with the right tools, wow. But I doubt it. It's too hard to hide with standard assumptions.

I would love to see the function for magnitude vs optical cross section, no doubt one can calculate that from first principals. It's just observable star density.

You need to determine velocity and distance in order to determine if the thing you just saw blink out a star is in orbit around the Earth (or not). There's a lot of distance between the Earth and the star, and literally anything between your eye and the star could obscure it from your perspective. We aren't even close to detecting/tracking all of the objects in the solar system that could do that. I'm also assuming the thing is not reflective (per the comments earlier I was originally replying to).

> An adversary can (presumably) almost immediately target that location if they are willing to admit they saw it

You'd be late to the party. If it was in orbit around the Earth, and since you don't know its direction or distance from Earth, your impact object would have to do a lot of searching once it got into space. If you had a giant laser to destroy it and could fire instantly upon detection, it will have likely moved quite a bit in the fractions of a second that it took the light to reach it. And, again you don't know the direction it is moving.., you can't compensate.

Ah, ya I was assuming if it blinked it is most likely in orbit around earth. Not sure how true that is, although far enough away similarly sized objects get diffracted around. I should try to math;) As for target I just meant light it up, properly diverging hp radar or laser pulse in an atmospheric window.

It does seem that it would be much much simpler to "hide" in another satellite, i.e., just have a third-party cutout launch a run-of-the-mill communications satellite and piggy-back your top-secret spy shit onboard. That way anyone looking will just see the communications satellite. No need for stealth

There's no way to keep that kind of thing secret. Too many people involved in building commercial satellites. Plus commercial comsats are either too small, or in the wrong orbit.

US intelligence agencies do purchase some imagery from commercial photo satellites. It's lower resolution than spy satellites but helps to fill in for coverage gaps.

I've seen WorldView-3 in person before it launched. The optics are enormous relative to the size of the also large bus. You couldn't hide that. Plus the changing attitude would point to an Earth observing mission.

This is especially sad as Grumman made the LM ("LEM" -- Apollo lunar lander)

Actually, looking at their publicity page, various companies and divisions that made parts of the Apollo program got rolled up into the current conglomerate.

Edit: forgot to include the publicity/brag page: http://www.northropgrumman.com/AboutUs/OurHeritage/Pages/Ins...

What would the condition of the satellite be if it re-entered the atmosphere? Total burn up? Or is there some remnants of a $3.5 billion dollar object that are worth recovery - at least to ensure none of it got picked up by another state?

The SpaceX's FCC application[1] for their broadband satellite constellation came up in a recent discussion here. Page 65 has tables describing the parts of the satellites that are expected to survive all the way to the surface. These are very different kind of satellites compared to Zuma, but the tables (and the text they accompany) might give some idea of what kind of components might survive (until impact, anyway).

[1] https://licensing.fcc.gov/myibfs/download.do?attachment_key=...

They wouldn't want anyone to be able to recover anything from a re-entered satellite, and would presumably design it so that nothing would survive reentry. At some point, at the end of its service life, they would likely have wanted to de-orbit it anyway.

I presume building a second would cost far less.

Don't companies build at least two versions anyway? There are three almost exact replicas of the Curiosity mars rover if I recall correctly. Building two copies instead of just one adds very little cost (relative to the overall cost of the whole project).

One JPL-related source [1] indicates that the second copy can generally be obtained for 15% of the price of the first one. The primary source is "Interview with Gentry Lee, Jet Propulsion Laboratory, June 1, 1984." I doubt things have changed very much since.

[1] https://history.nasa.gov/computers/Ch6-3.html

I toured JPL years ago and it sounded like the duplicate rovers were for testing maneuvers and software updates after launch. More for testing/debugging than a failsafe.

Non paywall version anywhere?

try opening the article and then paste this into your browser's URL bar:

  javascript:window.location.href='https://m.facebook.com/l.php?u='+encodeURIComponent(window.location.href);

$3.5 billion dollar satellite? That must've been one hell of a piece of equipment. I really can't even imagine would was on that thing that would cost that much to develop.

From my understanding developing Military Satellites is far more expensive than developing commercial satellites. The reason is actually the amount of secrecy and protocol behind the development of classified technologies.

Building a satellite is already an ultra-complex project that requires designing and sourcing specialized parts that need to comply with some very detailed specs.

I'm speculating but I'd have to imagine that a project so heavily classified, would result in a larger cost (reduced availability of contractors, more time required to design specialized parts, less availability of researchers and manufacturing workforce that meet the security criteria, etc.)

Just a reference, Envisat which was was an earth observation satellite developed by the EU, had a cost of around of $3 billion USD.

Hi.

It doesn't matter how he answers them, if he answered them in a way you could interpret.

Interesting - that statement kind of puts you both in danger.

Or he was trolling you.

I can only hope, I didn't like his answers, from a security perspective.

Ha! That’s awesome. :)

Most of the expense in government/military contracts comes from the many, many layers of "management" that have to be continuously fed a diet of large stacks of cash. Then you have to factor in the perverse incentives created for the contractor, where the longer they can drag a project out for and the worse of a job they can do and the cheaper the people they can hire to just barely get the sign-off for completion at the end, the bigger a success it is for the contractor. I worked for quite awhile doing maintenance on a system which was originally developed by an external contractor (one of the ones whose names appears in the comments on this thread). There was originally a contract to develop 3 separate systems. After years of delays and renegotiations and whatnot, 3 separate contracts had been given to the company, and once the government got sick of throwing good money after bad and said 'just give us whatever you have', they got 1 system delivered. That took 9 months of continuous effort to get into a state where it could be used. You know what you call that? A stellar success. At least, if you're the contracting company.

That is the program cost. Like with software, much of this cost would be in the design and intellectual property of the program. The replacement cost, the cost of goods sold, is far lower.

a lot of the cost of a given spacecraft is the years of test and verification that goes into inspecting the hardware and confirming that it is flight-ready. that involves dozens or hundreds of engineers working long hours in clean rooms with enormous vacuum chambers.

also god only knows what the electronics on these spy satellites looks like. i wouldn't be surprised if there is significant custom silicon on them.

Maybe the $3.5 billion dollar transaction was the product.

true, no one can criticize your product if you crash it into the ocean

Why the down votes? I think it’s cynical but certainly within the realm of possible.

>I think it’s cynical but certainly within the realm of possible.

Just about anything is "within the realm of possible". Without even a shred of actual positive evidence, this is just obnoxious flat-earth level conspiracy bullshit. You could make the exact same statement about just about every single government expenditure you are too lazy to research. Unless you have a basis for suspicion, this is just a waste of time that is pandering to people who flatter themselves with their "realistic" worldview.

Well, a lot of the discussion on this post is about other possible conspiracy theories (imho less cynical than this one), so I thought I’d was relevant that the GP was suggested as one possible option. The other theories are not much supported either.

Reply on news.ycombinator.com