The US Government is Finally Hitting CTRL+ALT+DEL on its Space Acquisition Strategy
Signs of rapid change emerge in recent large program cancellations and new space triumphs
TL;DR
This week we saw both DoD and NASA cancel two large multi-billion dollar Space programs, one a classified communications satellite being built by Northrop Grumman and another OSAM-1, an orbital services demonstration mission being built by Maxar. The other shoe has begun to drop in the industrial base as well, as we’ve seen the retooling of “old space” with Northrop laying off 1000 people in Space Park (Redondo Beach) and a rare RIF with JPL in Pasadena laying off 530 people due to slowdowns on planned deep space missions.
Some decades nothing happens, and some weeks decades happen. - Lenin
In sharp juxtaposition with old space, new space experienced several triumphs this week: Varda successfully recovered their pharmaceutical manufacturing capsule; Intuitive Machines landed on the moon, the first ever for a commercial company and the first US moon landing in 50 years. SpaceX launched their 8th manned mission to the ISS and Apex and True Anomaly’s first launches made it to orbit with widely awaited first satellites as well. So while old space is seeing major setbacks, new space is advancing at a dizzying clip.
We are in the thick of a force and industrial base transformation in space. The government space enterprise, long centered on a few multi-billion custom built exquisite systems manufactured by a handful of enormous contractors, is transforming to a more nimble and diverse contractor base focused on smaller, rapidly replaceable systems built with more standard components. This has been in works for quite a while and will by all indications lead us towards a new set of space assets which is more nimble, resilient against adversarial threats, and more rooted in a more diverse commercial industrial base of smaller suppliers, leveraging commercial electronics roadmaps that can evolve much more quickly than in the past.
I spent a large portion of my early and mid career working on space systems - mostly as a payload systems engineer. Systems like VIIRS, STSS, O3B mPower, APS Glory, PTES (and others I won’t get into). I have at least twenty payloads and missions on orbit I’ve worked on- which with the exception of the tiny bit of work I did on the Artemis I payload and my investment in Wyvern space - would almost entirely fall under the moniker of “Battlestar Galactica” programs- exquisite multi-billion dollar systems that took decades to put on-orbit and with expected lifetimes in the decades length as well. This broken acquisition model is quickly becoming history.
From the sidelines I’ve watched as the “old space” model has been overtaken by a move to smaller, more focused systems, “new space” with shorter lifespans. It became obvious this week that this change has become an avalanche. Having a foot planted in both worlds, I have a unique perspective on how we got here and where we may be going.
In this piece, I will summarize the history of the American space industry and how we went from a country that put a man on the moon in nine years during the sixties from a standing start to an enterprise that hit rock bottom in the early 2000s as a risk averse, mission assurance culture at NASA and DOD left us bleeding red all over the budget on huge satellite programs; dependent on the Russians for both engines for our satellite launches and ferrying our astronauts to the international space station. I will then talk about what changed in the last few years and where I think this ends up.
The Right Stuff
“For a test pilot the right stuff in the prayer department was not “Please, God, don’t let me blow up.” No, the supplication at such a moment was “Please, dear God, don’t let me fuck up.”
― Tom Wolfe, The Right Stuff
As anyone who has watched The Right Stuff can tell you the original astronauts, known as the Mercury Seven, were cut from a different cloth. These men were test pilots (and most were air combat veterans) in an era where the average Naval aviator had a 23% probability of dying in an accident in a 20 year career (that’s not including combat). Aviation, let alone space exploration, was not for the faint of heart.
To compliment these test pilots, NASA’s engineering culture in the early 1960s looked a lot like SpaceX’s today: not afraid to take prudent risks, fail forward and guided by extraordinary pragmatism and common sense of purpose. The DNA of the organization was drawn from a combination of these test pilots, along with GI bill educated engineers from the Greatest Generation, along with ex-Nazi rocket scientists like Werner Von Braun and others Operation Paperclip. Can you imagine the political fallout today if an ex-Nazi were to be head of NASA like Karl Debus was in the early 60s?! This was a no bullshit organization that was extremely mission driven. The DNA was driven by a desire to “fail forward” and the theme of the day as they put in 100+ hour work weeks was “waste anything but time.” Their attitude and swagger in an era of heroic engineering was best epitomized by Captain Kirk’s soliloquy about “Risk is our business”:
NASA and the legacy of 17 dead astronauts
"In 1966, NASA took over in space, and it has been a bureaucratic mess ever since." - Chuck Yeager
Then NASA had its first of several tragedies and the political blowback made them retrench and get bureaucratic. Three astronauts were incinerated in a fire on the landing pad during ground rehearsals for Apollo 1 in 1967. This led to Congress to lean hard into NASA, including the infamous Phillips Report, a memo about NASA safety oversights earlier in the program that wasn’t even related to the root cause of the Apollo 1 fire, but was a convenient prop for (then Senator) Walter Mondale to wave in front of NASA as a sign of reckless behavior. This caused a shift in NASA’s playbook to be more cautious. However, the slowdown was not enough to stop us from getting to the moon six times, getting Skylab on orbit, launching numerous missions to Mars and the outer solar system and starting the shuttle program.
By 1986, NASA had gotten it’s mojo back: there was talk of a West Coast shuttle program from Vandenberg, of telescopes on orbit, of returning to the moon and going to mars, and a Space Station on-orbit by the early 90s. Riding the coat tails of the defense buildup under the Reagan administration, NASA was again aggressively pushing forward. Then disaster struck at the worst possible time: with millions of schoolchildren watching from their classrooms as the first schoolteacher, Christa McAuliffe went into space, the Challenger blew up 73 seconds after leaving the pad due to a cheap rubber O-ring on one of the solid rocket boosters getting a crack because it was operating outside its designed temperature range. NASA literally had failed in front of the entire nation and as the Rogers commission later called it a “failure of history” with echos of the Philips report many years before. NASA curtailed its use of the space shuttle for any military missions, stopped carrying commercial satellites and took a 32 month break between launches. They were back in the safety penalty box.
17 years later, NASA was finally getting its groove back. Nine years of leadership under Dan Goldin and Faster, Better, Cheaper had given us multiple deep space missions, 75 shuttle flights, an International Space Station (ISS) being assembled and the Hubble Space Telescope which we were able to fix on orbit. Then disaster once again struck: a piece of foam broke off the external fuel tank and cracked a few thermal insulating tiles on the bottom of Columbia. The space crack disintegrated in the hot plasma of the atmosphere while returning to earth. Once again a board was commissioned (the Columbia Accident Investigation Board) and once again the same cultural problems were uncovered. Space Shuttle Program Manager Wayne Hale put it quite succinctly: “As in every accident investigation, the investigation board found that communications between people and organizations were faulty. Management culture was poor. And the safety organization was strangely silent on dangerous situations which they had been warned about. So the recommendations, in additional to technical things, included improving communications, changing management culture, and reinvigorating the safety organization”
“ So the recommendations, in additional to technical things, included improving communications, changing management culture, and reinvigorating the safety organization” - Wayne Hale, former NASA Space Shuttle Program Manager
NASA was a victim of its own failures and every subsequent mission, be it manned or unmanned was plagued with the fear of failure and the ensuing cost spikes as a result. This resulted in bloated requirements creep, unnecessary inspections and a general paranoia about reliability that slowed everything down. The scar tissue of the 17 dead astronauts was impeding the range of motion of the whole organization.
Contractor Base Consolidation
While NASA was dealing with the scar tissue of 17 dead astronauts, the fallout of the peace dividend and a radical contraction of the industrial base from 75 prime contractors down to 6 was also underway. What had started in the 1980s due to a combination of financial arbitrage maneuvering going after overfunded pension funds and the breakup and sale of Hughes to Boeing and Raytheon due to attacks on it’s non-profit status kicked into overdrive after the so-called “Last Supper” in 1993 which Defense Secretary Bill Perry hosted at the Pentagon where contractors were actively encouraged if not coerced to consolidate, as Figure 3 shows.
The withering of the industrial base led to a massive brain drain and reduction of competition, removing natural pressure for efficiency and cost reduction. The increasing prevalence of cost-plus contracts and the reduced competition made the remaining primes standing larger and driving up their market cap, as shown in Figure 4.
So, less competition coupled with a culture focused on reliability. What could go wrong?
US launch platforms become dependent on Russian made rockets
Something else shocking also happened in this time: the US launch enterprise, through its own conservatism and consolidation, became entirely dependent on Russian made engines to get into orbit, in particular the RD-180 which was used for the Atlas V rocket. In 2006, we consolidated to a single launch provider capable of handling large national security satellites to LEO and GEO, the joint venture between Lockheed and Boeing known as the United Launch Alliance (ULA).
Our dependence on the RD-180 for space launches from about 1999 to 2015 when SpaceX managed to break the wheel and break into DoD launches was an unintended consequence of the last supper defense consolidation mentioned above (GD-Convair built the original engines for the Atlas V; they were bought by Martin Marrietta in 1994; Martin Marrietta is now part of Lockheed Martin), corporate expediency (the RD-180 was licensed as a gap filler that never got filled), and irrational conservatism (“if it works and its approved and proven, don’t replace it”) brought about the mission assurance regime of Old Space, which I’ll get into next.
Indeed, its really only been since the Russian invasion of Crimea in 2014 with the Russians telling us they flat out won’t sell us the engines anymore that we’ve seen a concerted push to find alternatives to engines for the Atlas V (SpaceX and others of course have rockets that don’t need Russian engine).
Non-virtuous cycle of Mission Assurance
In the late 90s, the buzzwords “mission assurance” came into vogue across the entire government space enterprise. While NASA remained the most extreme example, the push for Mission Assurance began to really take off across the entire government space enterprise, partially as a result of a string of launch failures resulting in the loss of five Air Force and classified spacecraft.
The mission assurance initiatives started with the laudable goal of reducing failures and creating a culture of engineering rigor. Unfortunately, its end result was creating a culture with an irrational emphasis on quality. Many of the aerospace contractors added legions of watchers in their quality and process organizations that were incentivized to stop progress for any reason. Adding further incentives to slow work was the prevalence of cost-plus contracting, which encouraged spacecraft to stay in the high-bay as long as possible without the contract getting cancelled where they could collect engineering change proposals that would generate more fee.
Some of these so-called “mission assurance” requirements approached lunacy. I recall hearing a story of NASA sending a government employee out to antenna subcontractors in Europe to physically observe every solder joint being made on every single one of the 50+ antenna elements in a payload- something which industrial process controls like AS-9100 and J-STD soldering standards is supposed to handle (ie. certify the production line). This is like demanding that you personally inspect every grain of wheat in your bread loaf to ensure it isn’t contaminated rather than just taking the FDA inspector’s word for it.
For the first 7-9 years of my career, the NASA story was closer to the rule than the exception. No wonder we couldn’t build anything!
You get what you incentivize. Having watched this feedback loop in action, the parallels to how DEI spread and dug itself in as its own non-value added bureaucracy across corporate America a decade or so later aren’t lost on me. With contractors incentivized to encourage more and more reliability, and the government encouraged to not led suppliers handle their own quality but instead outsource it to independent organizations who made money off finding problems and prevent product from shipping, and everyone gets more and more cost plus incentive fee as the contract grows - what could possibly go wrong?
This led to the creation to a whole cottage industry of mission assurance bureaucrats working for not just companies internally but with FFRDCs like the Aerospace Corporation (“circle A”) and SETAs (Systems Engineering Technical Assistance) contractors like Jacobs, Odyssey, Matrix, Centauri, etc using this as an opportunity to increase their bottom lines (see Figure 6).
The negative feedback loop created, compounded by our scarcity of reliable launch options which made every mission just that much more important, led to a death spiral as seen in Figure 7 below. Old Space did everything it could to keep this going and the end result was dozens of Nunn-McCurdy breaches (where the program exceeds its budget by 15% or more) on critical programs like SBIRS-HIGH, NPOESS and others that brought the whole space enterprise to a grinding halt. Figure 7 shows this non-virtuous cycle in action.
Enter Elon Musk - SpaceX and How We Stopped the Death Spiral
There was a better way, more culturally rooted in the fail-forward, take risks approach of the early days of space travel. This was the Silicon Valley, agile development mindset that Elon Musk and his team at SpaceX brought to new space starting in 2002. The biggest difference between new space and old space is rooted in the idea that if you aren’t at least failing occasionally, you probably aren’t going fast enough or at least taking reasonable risks.
The biggest contrast between old space and new space is the idea that if you aren’t failing at least occasionally, you aren’t going fast enough or taking reasonable risks
The other factor that Elon Musk brought to the table with SpaceX was a set of heuristics that accelerate the design cycle and time to market that are in many ways anathema to old space, with its cost plus and mission assurance incentives, that exist today.
His five core rules that he spells out in the video above I can literally provide an example in my experience in old of how we did the opposite for everyone of his core points:
Make you requirements less dumb
Delete the part or process
Simplify/optimise
Accelerate cycle times
Automate
Old space programs to this day are often based on a “thin prime” model: Northrop, Boeing and Lockheed would provide the bus and a cadre of smaller subcontractors would design the payloads and components.
Ironically, the only times that Old Space actually works relatively well is when it is vertically integrated, like in some of Boeing’s communication satellite programs. The inability to build things in house leads to a complex web of contractors, with indirect and direct oversight, with a similarly complex set of politically defined interfaces, which causes delays to cycle time and proliferation of translation errors between contractors which often led to random system defects. You end up with complex interfaces where in a vertically integrated supplier none would make sense. A key lesson I learned early in my career as a systems engineer is that the real elegance of an architecture isn’t in the requirements, its in the interfaces.
A key insight I observe early in systems engineering career is that the real elegance of an architecture isn’t in it’s requirements, its in the interfaces.
New space often takes the opposite approach: favoring vertical integration and following Elon’s algorithm, questioning every feature and looking for the absolute simplest way to get the job done. Old space favored, big systems like the cancelled NPOESS program I worked which on had as many as 20 instruments on an enormous spacecraft. New space prefers small spacecraft with single functions or few functions and shorter lifetimes. The benefit is faster execution time, dramatically lower cost and more shots on goal from a mission completion standpoint.
Rideshare, ESPA Orbital Transfer Vehicles and the dramatic drop in cost
Some other innovation have occurred on the launch side that has really brought price down in the last five years. Foremost amongst those has been the introduction of rideshare as a business model by first SpaceX in about 2018 then next Rocketlab, and now eventually pretty much every launch provider. With smaller satellites becoming more prevalent due to more focused missions and shrinking electronics, rockets have gone from being Ubers to orbit to buses and the cadence of launch has risen dramatically. As figure one above shows, its not extremely common for large rockets like SpaceX Falcon 9s to carry 50 or more satellites from many different customers into orbit. This would have been unheard of 15 or even 10 years ago.
Another factor, perhaps as enabling for the new satellite industry as the introduction of the standard sized shipping container was for maritime and railroad logistics in the 1950s and 60s, was the introduction of EELV secondary adaptor payload (ESPA) ports becoming ubiquitous by 2015. Suddenly small satellites could be built to generic standards rather than having to be custom built and certified for particular launch platforms. This enabled launch platform agnostic buses and hosted payload business models such as those used by Loft Orbital, Clyde Space and others to also be possible: they can just book space on a rocket and then fill up their buses with payloads from third party providers.
Rideshare has turned rockets into the space equivalent of container ships, ESPA ports have turned smallsats into shipping containers
Making things even easier, we are now seeing the entry into the market of orbital transfer vehicles (OTVs), with providers such as Momentus, Voyager and 43 others offering services to put your satellite into the precise orbit after launch that you want. This enables a single rocket launch to support even more missions. The rocket gets you to the right train station, then you take an OTV Uber to your final orbital destination.
Bottom line: it’s never been easier to build space craft and launch them to exactly your orbit as the standardization and infrastructure, in addition to just the economies of scale have finally gotten there.
Disaggregation and the SDA
I recall first hearing the word “Disaggregation” in a space context nearly a decade ago in a speech from General John Hyten, then head of US Space Command. It was a cornerstone of a larger strategy known as “Resilience”, which was the recognition that space was a warfighting domain and that we needed to take steps to defense ourselves from attack there. It stemmed from the realization that while our exquisite space based assets gave us an enormous advantage versus our adversaries, their huge cost and their near term irreplaceability made them very vulnerable in a near-peer adversary conflict. This was made all too clear by Russian and Chinese displays of ASAT capabilities like the 2007 ASAT missile test against a defunct Chinese weather satellite (where I first was exposed to the Space Control Domain).
The Space Development Agency was formed in 2019 with this idea in mind and drew inspiration from new space with a high emphasis on standardization, building large constellations of small sats in primarily low earth orbit to do jobs traditionally done by a few large expensive assets, often in geosynchronous orbit (GEO). They are presently at 33 satellites and counting on orbit with plans to deploy a full constellation of over 200 satellites in multiple tranches over the next few years. One notable thing about them is the ease with which they select from a combination of traditional and non-traditional contractors for their constellations, switching between stalwarts like Lockheed, Northrop, RTX and L3 Harris as well as newer players like SpaceX, Sierra Space, Rocketlab, York Space Systems. This is giving the industrial base desperately needed supplier diversity.
History may in the long run show that it’s current director, Derek Tournear, is on the same level of effectiveness as a defense bureaucracy change agent as Bill Perry was (who gave us the Second Offset Strategy in the 70s and 80s - the tech transformation towards network centric systems, stealth and precision guided weapons that won us the first Gulf War). Hopefully he avoids the Last Supper trap that Perry fell into later in his career!
Geosynchronous orbits falling out of favor versus Low Earth Orbits
Another trend going on which is throwing old space for a loop: Mega-constellations of cheap rapidly reconstitutable assets - such as Starlink - are eating the satellite communications (SatCom) market traditionally dominated by GEO satellites.
Gone are the days of Terabit capacity satellites in GEO with 13 year lifetimes, enormous deployable antennas and 10+ kw of power capacity. They are being replaced by massive constellations of hundreds or thousands of small satellites.
Gone are the days of Terabit capacity satellites in GEO that require 10+ kilowatts of power and enormous deployable antennas - they are being replaced by the likes of Starlink, Kuiper, Intelsat, 03B Mpower, OneWeb and others with smaller, more affordable satellites in lower orbits
. This has put traditional GEO satellite powerhouses like Boeing Satellite Systems (BSS) in El Segundo, Maxar (formerly Loral) in Palo Alto, Airbus in France, and Lockheed in Sunnyvale at a crossroads. Traditional players were caught flat footed by this, often without robust smallsat offerings of their own as they had spent decades working on GEO bus designs that weren’t easily transferrable to LEO designs and made little economic sense. GEO orders are down 50% year over year this decade (see Figure 9) and price pressure will continue to make them less attractive as Starlink and other LEO constellations drive more competition.
As the LEO constellations enter full go to market stride, we are seeing them eating the market share from GEO providers, leading to consolidation. Viasat recently bought INMARSAT and more consolidations are likely coming. Adding further pressure, broadcast players like DirecTV (now AT&T) have now almost entirely switched over to 5G or using other services, so aside from earth observation and some critical government communication, I think this market is likely going the way of the dodo in the long term.
The Future
"It is difficult to make predictions, especially about the future.” - Yogi Berra
I’ve gotten myself in trouble trying to predict future trends before - for instance, five years ago I thought that Orbital Transfer Vehicles weren’t going to go anywhere when someone first presented me Momentus (while I was right about the company being set up for problems, I was wrong on the market). Turns out I was missing how rideshare was changing the market. However, I feel comfortable making a couple of simple prognostications that I’ll put forward:
SDA will continue to cannibalize capabilities from other agencies and we will see more large program constellations, some of which will occur away from plane sight (like we saw this week).
The GEO market will continue to contract to the point where it’s basically just legacy players and government.
The primes will need to put their balance sheet to work retooling their products through M&A, like Lockheed did this past week with its purchase offer for Terran Orbital this week, to survive.
Organizations rooted in providing “mission assurance services”, SETA services and other products will pivot away from this as the amount of money tied up in individual assets makes them less relevant. We are already seeing this with players like Leidos, Booz Allen Hamilton, CACI and others using their venture arms and M&A capabilities to buy small product companies to diversify away from just “rooms full of bodies”.
We could even “defense as a service” in space being given a try with the government outsourcing even orbital defense to companies like True Anomaly. Indeed companies like Hawkeye360, Vannevar Labs, LeoLabs and Slingshot Aerospace are already doing this to some extent or another.
The launch enterprise will overcome any last bottlenecks as Starship comes online and a glut of launch providers may drive launch prices down even further. As my friend and colleague, Liz Stein wrote about a couple years ago, some key bottlenecks in engine manufacturing (which is now mostly done with 3D metal printing in large beds) in terms of both size and volume are being overcome which will make Rideshare even more ubiquitous, increasing the demand for OTVs and other vehicles to support the “last-mile delivery” of these large constellations.
Some additional legacy players will get consolidated or their space divisions will get spun out to private equity and new players like Sierra Space, York Space Systems and Rocketlab will rise to take their place (SpaceX has already done this).
All of this leads to a future in which the old space world, where we used to joke that “when the paperwork weighs more than the payload, you can launch” is being replaced by a more agile industry able to counter America’s threats- the government to it’s credit is leaning in and recent program cancellations offer a glimpse into that thinking.
“All truth passes through three stages: First, it is ridiculed; Second, it is violently opposed; and Third, it is accepted as self-evident.”- Arthur Schopenhauer
So, while a decade ago it seemed like new space was having trouble getting off the ground and at risk, it’s now possible to see that it’s eating old space’s lunch and may consume it. It’s like that Schopenhauer quote about new ideas: “All truth passes through three stages: First, it is ridiculed; Second, it is violently opposed; and Third, it is accepted as self-evident.”