Expert Voices

When it comes to space, failure isn't just an option — it's a requirement

SpaceX's SN8 Starship prototype launches on a high-altitude test flight on Dec. 9, 2020. The vehicle performed well during the flight but did not stick its landing, exploding shortly after hitting the ground.
SpaceX's SN8 Starship prototype launches on a high-altitude test flight on Dec. 9, 2020. The vehicle performed well during the flight but did not stick its landing, exploding shortly after hitting the ground. (Image credit: SpaceX)

Sometime next month, SpaceX will likely try to launch the largest rocketship ever flown, known as the Starship. If successful, this flight will herald a new space age, as humanity opens the solar system. This begs the question: Will it fail or will it succeed? And if it fails, what will it mean?

Anyone familiar with space movies, or who has visited a NASA gift shop, knows the phrase "Failure is not an option." Actually, it is. In truth, if one is to win at anything, failure is a requirement. We only learn by what we learn from making mistakes. That is if — and this is an important "if" — we accept the learning we gain from the failure and apply it to our future endeavors.

The exploration and development of space are no exceptions. Or rather, they highlight the rule.

Related: Spaceflight: Danger at every phase (infographic)

Generally speaking (and in my humble opinion) there are four types of failures when launching a rocket or rocketship. 

  • 1. A new, untested tested technology or group of technologies fail. 
  • 2. An incorrectly tested technology or group of technologies fail.
  • 3. Humans in the loop ignore rules or design criteria from their own plans or cut corners, usually for financial, performance or political reasons.
  • 4. Having had one of the three types of failures above, no learning has occurred, or learning did occur and was ignored, and they happen again.

While any "rapid unscheduled disassembly" (RUD as some engineers jokingly call a rocket explosion) is a problem, number 1 in my list is an expected part of the development of any new complex system. Sadly, and by far, it is the second, third and fourth failure categories that most often result in tragedy. The two most spectacular American instances of this were the space shuttle Challenger and Columbia disasters. In both cases, the failures were predictable, if not specifically identifiable. 

Like all NASA human spaceflight programs, the space shuttle program was a rock star show of enormous proportions and massive implications for the agency and the government. In both cases, the inconvenient truths of possible deadly mishaps or results were spotlighted before the accidents by members of the mission management teams. While the case of the Columbia disaster is more complex and nuanced, in the case of Challenger, clear warnings from savvy employees were ignored because the show had to go on, and, sadly, people died.

Due to fear of such high-profile failures, our space agency usually over-corrects, or perhaps more correctly, over-corrects for the wrong reasons, and in the wrong way, as every tiny malfunction, mishap and mistake becomes the fodder of media pundits and agenda-driven critics. Witness the NASA Space Launch System (SLS) — as it was rolled out to the launch pad, rolled back, rolled out and rolled back, as engineers worked to assure everything was perfect — even as the wait, the holds and the rollbacks themselves actually added potential risk to the system. 

In the end, the politically designed contraption flew — successfully, on the Artemis 1 mission, but if this is to be the vehicle that opens the solar system, such fragility means none of us are going any time soon — if ever. It's all just too precious. Yes, I was one of those critics. I still am. And while an in-depth discussion of the disastrous pork program I've long called the Senate Launch Scam must wait for another time, it is clear that even as the space agency tries to rationalize one of the most expensive and complex rockets ever built, it is ironically hostage to its fear of failure. Failure made more likely by the very system that created the vehicle in the first place.

Related: The 10 greatest images from NASA's Artemis 1 moon mission

NASA’s Space Launch System rocket carrying the Orion spacecraft launches on the Artemis I flight test, Wednesday, Nov. 16, 2022, from Launch Complex 39B at NASA’s Kennedy Space Center in Florida.

NASA’s Space Launch System rocket carrying the Orion spacecraft launches on the Artemis 1 flight test on Nov. 16, 2022, from Launch Complex 39B at NASA’s Kennedy Space Center in Florida. (Image credit: NASA/Joel Kowsky)

This is no one's fault. It is a result of a systemic fault created over time. Today, especially in the government, too much rides on every liftoff. The weight of fear of failure means that it will not be gravity that stops the human breakout into the universe, it will be bureaucracy, and a system that has gone off its tracks and rolled off the playing field into a swamp of its own making.

But it wasn't always that way. Starting after World War II, as missiles previously designed to kill people in wars and tested under the military's cloak of secrecy began to be employed in exploration, experts understood that, when testing new tech, failure wasn't just an option. It was the rule — until the cameras showed up.

A good example occurred in the late 1950s, as the massive global science project known as the International Geophysical Year (IGY) was getting underway. Launching shiny objects into space was an obvious yet new exploration arena, and those in charge adopted fairly pragmatic expectations. In the U.S., the National Academy of Sciences agreed that if only one of every six satellites in the program made it into space, it would be considered successful. Flying on the Air Force's Vanguard rocket, three of them did. Yet to many in Washington and the public, the program was considered a failure.

Meanwhile, the Soviets were driving hard, and while they had their own fair share of failures, some of them spectacular, the world would only hear of those that succeeded. To be clear, Soviet, Russian and Chinese space failures deserve their own essays and volumes, and in many cases make the causes and results of U.S. failures seem small by comparison. This is in some ways the point. When you allow the world to watch, and give taxpayers a veto based on what they see, the pressure to perform becomes a force of its own.

For example, under pressure from the Soviet Union's Sputnik surprise, from 1958 to 1960 NASA tried eight times to launch Pioneer satellites to the moon using early variations of the Atlas rocket. One probe was never launched. The others all failed. Only Pioneer 5 could be called a success. 

Thus, NASA was born in the spotlight. And it was forced to either entertain the taxpayers for its funding or be dragged off the financial stage. Failure, even expected failure, even failure followed by success, became less and less of an option. It just didn't make for good TV, and bad TV shows get … well, they get canceled.

NASA's official history website is surprisingly frank in its assessment of this reality (italics mine):

"(The Atlas failures) made plain that success on space missions would be neither automatic nor cheap. There was a price to pay, and part of that price was the failure of some missions. This price NASA management would never find comfortable. The trouble with a philosophy of accepting a certain number of failures as normal and inevitable was that even 'learning failures' in an open program like NASA's, conducted under the watchful eye of a whole world, looked to the public and Congress like absolute failures. The press treated them as failures. It didn't matter that previous development projects like the V-2, Atlas, Thor, Polaris and almost any other major rocket development one could name, had had their share of unsuccessful early firings, and that this had been accepted as a necessary growing pain. Those difficulties had been hardly visible under the cloak of military secrecy. But space program difficulties were highly visible and distressing."

"Distressing." An understatement, to be sure. It became clear to supporters of NASA that, in many ways, our space program wasn't just about exploration or strategic victory over our communist opponents. It was entertainment. Thus, the show must go on. Whether it meant creating massively over-engineered systems, massively over-papered bureaucracies or both, NASA had to be damn sure that whenever the eyes of the world were on it, its people would deliver a Right Stuff performance.

Thanks to an incredible level of proficiency on the part of NASA's team, for the most part, this approach worked during the Apollo program. Yes, there were terrible disasters like the Apollo 1 fire, which killed three of the agency's top young astronauts. But that disaster happened during practice, and no one was watching. It slowed down the program, as NASA reviewed its mistakes, and made serious improvements in its systems and procedures. But had the fatal fire happened live on screen, the entire program would most likely have slipped years into the future as the various branches of the government each took their turn at reviewing how such heroes could have died — and quite possibly, given the horrible nature of their death in a burning capsule, it might have ended the entire program.

Related: NASA's 17 Apollo moon missions in pictures

What is remarkable about Apollo is that there were so few significant failures. Think about it. NASA went from almost zero spaceflight experience to humans on the moon in less than 10 years, with only three people dead in one major accident — which didn't occur during a mission. Yes, there were spacecraft failures and blow-ups, but history has already correctly rewarded the entire program with a well-deserved mark of excellence. Ever hear the term "Moon shot?"

Then, as I mentioned, came the shuttle program, with an entirely different system and an almost all-new management team, given the time lag between the end of Apollo in 1972 and the first flight of Columbia in 1981. Ironically, the space shuttle was supposed to be the system that would make human access to space safe and routine. It was almost the opposite. To get it funded, the original plan was compromised to save money and gain support. For example, as proposed to Congress, the shuttle system consisted of two totally reusable flyback systems that would take off vertically and land on runways, ready to fly again. As political sausage-making took over, however, this approach was replaced by the massive hydrogen and oxygen tanks and solid rocket boosters we are familiar with today. 

As is usual when trying to win votes, political games took precedence over practicality. Worse, while trying to deal with the political compromises being forced on it to get the funding it needed for its dream ship, NASA layered even more bad choices into the mix. Here's one example of many cited in the many retrospective books on the disaster: To win votes from the then-powerful Utah congressional delegation, military intercontinental ballistic missile (ICBM) solid rocket manufacturer Morton Thiokol in Utah was chosen to manufacture solid rocket booster segments that had to fit perfectly together. These were then shipped to the Cape in Florida on their sides on train cars over hundreds of miles of railroads. 

In the end, the shuttle was only partially reusable, and became so complex that the entire vehicle had to be torn apart and rebuilt between flights. It was, according even to NASA leadership, a recipe for disaster. After Challenger, former NASA Administrator Norm Augustine mentioned on a conference call his concern for a program based on "flying space shuttles in circles until another one falls out of the sky." And it did. And it was Columbia.

So now the shuttle program is gone. Frankly, this is a good thing, in spite of all the good it did to inspire a generation. I helped make it go away. It was too expensive and dangerous, and rather than helping open the space frontier, it was helping keep it closed. Once the quasi-military propaganda campaign to win the race to the moon was won by Apollo, our government should have switched to commercial fleets to take over basic shipping and people-carrying to and from space, just as we do in the rest of our economy. A group of us in the Space Frontier Foundation spent over a decade making that happen. Eventually, at least to and from the International Space Station, we won.

This begs the question: Are private space transportation systems safer? Are private operators more careful than government agency contractors? Good questions. The first answer is that no engineer or project manager ever wants anyone to get hurt — no matter who they work for. But systems don't care, and systems designed around an internal conflict between the need to break things to learn things, and not being able to break things because it doesn't look good, are inherently dangerous.

At the end of the day, blowing up very expensive hardware is not fun for anyone, government or commercial. Unfortunately, sometimes one has to blow up a few things to get things right, and the way our culture is wired, the private sector has more freedom to do so than NASA right now — at least the part of the sector with enough money to blow up stuff and replace it.

And so, here we are. In the coming weeks, Elon Musk, the man who has blown up more big rockets than anyone in the last 10 years, will try to fly the biggest rocketship of all time, the Starship. But, if you recall my opening list of why rockets blow up, Mr. Musk has focused on #1 — blowing up new and untested rockets to learn how to make them better. Better yet, he and his team are actually applying — rather than ignoring — their data to make the next rocketship safer! This is where he and other NewSpace companies have a distinct advantage over their government competitors. Their process is complete when they deliver a highly reliable, low-cost, reusable space transportation system that actually works. 

A good way to spot a technology-edge crossing company that is a bad investment is one that does not have this built into their budget, either through overconfidence or naivete. A potentially successful company combines vision with pragmatic pessimism. They err on the side of expecting early failure, and their budget includes funding reserved to try again until they get it right. Thus, to these companies the optics of failures do not matter as much as the application of lessons learned by those failures. Be it satellite companies, government customers or, eventually, citizen space pioneers, there is no inherent conflict in the success metric for a private commercial space transportation provider. It delivers, or it does not. The market handles the rest.

And so, as critics and competitors wail at his failures, Elon has rolled out, blown up, rolled out and blown up half a dozen test vehicles to get to the point he and his SpaceX team are at today.

So what about Starship and the Super Heavy monster first stage on which it will ride? Well, we know Starship has been vetted, broken and vetted again, but Super Heavy has not. Yes, SpaceX has fired the engines many times. And in some cases, they have not worked perfectly, so engineeers made some tweaks and fired them again. But will they all work when needed? Will all of the myriad of systems work that are required in order to push that shiny but heavy-as-hell rocketship sitting on top of it all the way to a successful splashdown or landing? We shall see.

What we do know is that, if it doesn't work, two things will be true: 1. Unless there is some strange anomaly in its trajectory if it fails, no one will likely die. 2. If it does fail, you can bet SpaceX will find out what went wrong, fix it, roll out another one and fly it again. And from what I know of Elon, Gwynne Shotwell (SpaceX's president and chief operating officer, who's running the Starship launch program) and the team at SpaceX, they will keep trying until they succeed. But they won't stop there. According to their plans, Starship, once proven, will be flown dozens, if not hundreds of times, carrying satellites and other commercial cargo into space before a single human life is put at risk in one of the hundreds of available seats they will screw into the floor once those flights prove the system is safe.

Related: Starship and Super Heavy: SpaceX's Mars-colonizing transportation system

A SpaceX Falcon 9 rocket launches 21 Starlink "V2 mini" satellites from Cape Canaveral Space Force Station in Florida on Feb. 27, 2023. (Image credit: SpaceX via Twitter)

This is part of the genius of Musk and SpaceX's plan. They have created their first customer base in the form of their own Starlink internet satellites, which can launch in the thousands on Starship. These relatively tiny revenue-producing robotic phone lines in the sky will fund SpaceX's development program and safety certification. Their own products will pay for the development of their own product as that product makes those products cheaper to use and vice versa. (It's brilliant, but I don't want Elon to know that, lest it goes to his head.) Thus, Musk can get his transportation system operating by risking his own money, then transition to outside customers who want their missions flown at less risk and then, in the long run, drive the risk down so low it will make sense for those first brave human pioneers to climb aboard and head off to the moon and eventually Mars. (Not to oversimplify, as each of those destinations will require yet another series of capabilities and the development of new systems.)

Contrast this with the broken system at NASA. Having just flown a significantly delayed first flight of an empty spacecraft on Artemis 1, it will be nearly another two years before the next flight, Artemis 2, attempts to carry a crew around the moon in a redux of the flight of Apollo 8 over 55 years ago, using a system — the SLS rocket and Orion capsule — that is essentially the same as those developed 70 years ago. One might expect it to be cheaper and more reliable given its heritage, but no, the opposite is true. At an estimated $4 billion per flight, the SLS is so expensive and complex that it will only fly every couple of years. This pace assures a stale learning curve and the turnover of experienced team members during the program's lifetime. Meanwhile, with the cameras of the world (and, more importantly, U.S. taxpayers) on them, each crewed SLS flight becomes so precious, once again, that failure will not be an option — within a system that increases the probability that it will occur. 

Meanwhile, Musk states that Starship will eventually cost just a couple of million dollars a flight. I know too many engineers to believe such a number, so let's make it 10 times as much, $20 million a flight. That's still massively cheaper than SLS. On top of this, Starship will be racking up dozens of flights a year carrying expendable mobile phone systems and other commercial and (interestingly) NASA payloads. Some may fail, technologies will be replaced and upgraded, but the flight teams will become even more honed — before the first human payload enters the airlock.

Hopefully, the SLS will become obsolete before it is called on to try to carry too many astronauts into space. This is for the simple and obvious reason that all the government paperwork in the world, all the covering of one's behind and all the going slow and being afraid of breaking anything will not make up for a broken system that has lost the ability to build things, break things and then build them better, something our government used to understand and seems to have unlearned.

So, when you watch the upcoming first test flight of Starship and Super Heavy, understand it is a test. It is not the Super Bowl or World Cup. It is practice for the Super Bowl. It is a rehearsal for the World Cup. If SpaceX fails, they will learn, just as the other new commercial spaceship companies are learning. If they succeed, they will also learn; soon enough, they will fail anyway. Yet the difference will be that the necessity of that failure is understood, the minds to learn from it are open to the knowledge it will provide, the risks next time will be mitigated, and the commitment to getting back up and trying again is built into the game plan.

The difference between a winner and a loser is that the winner gets back up, learns from what they did wrong and does it again.

This is how we win the future.

Follow all of the Expert Voices issues and debates — and become part of the discussion — on Facebook and Twitter. The views expressed are those of the author and do not necessarily reflect the views of the publisher.

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at:

Rick Tumlinson
Founder of SpaceFund, Earthlight Foundation, Space Frontier Foundation and New Worlds Institute

Both a rebel and a respected leader, Rick is listed as one of the top 100 influential people in the space field. Called one of the world’s top space “visionaries,” Rick helped coin the term "NewSpace" and worked to create the new commercial space industry highlighted by Elon Musk and Jeff Bezos.


A leading writer, speaker, and six-time Congressional witness, Rick helped start the first mission to find water on the Moon, signed the first ever commercial data purchase agreement with NASA, helped start NASA’s Lunar Exploration Analysis Group, led the commercial takeover of the Russian Mir space station, signed the first private astronaut to fly to the space station, co-founded the Space Frontier Foundation, and was a founding board member of the X-Prize. 


As a result of his world-changing work, in 2015 he won the World Technology Award along with Craig Venter of the Human Genome project. He founded the SpaceFund venture capital company with 19 space companies in its portfolio and is a member of the US Space Force Doctrine Advisory Group.


Rick's The Space Revolution "radiopod" is featured on IHeart Radio Network's iRoc Space Radio and is available on most major podcast sites. He hosts the New Worlds Conference and the Space Cowboy Ball in Austin, Texas, and his non-profit EarthLight Foundation is creating an inclusive new movement to use space to protect the Earth and expand life into the cosmos.

  • Pogo
    That’s pretty much how everything we do works. If it breaks, fix it or build a new one, but make it better. We learn from mistakes and stuff that didn’t work right. We walk better at 14 years of age than 14 months. 30 year olds drive better than 16 year olds (the 16 year old would disagree). Cars, airplanes, buildings, even hand tools work better today than before due to better design after previous ones broke.
    Most U.S. OSHA regulations are the results of accidents, not because they foresaw a problem.