Half a century after the flight of Yuri Gagarin awakened humanity to the high frontier, crewed spaceflight is still enormously expensive and dangerous. We remain in the same basic era of capability as those early pioneers, only with greater technical competence at using more or less the same technology - and yet paradoxically with far less confidence or clarity of vision in applying what has been learned. This, which we can call the First Space Age, has dragged on far longer than anyone had hoped or predicted, and kept both the cost and risk of human spaceflight at unacceptably high levels in pursuit of increasingly petty and unambitious objectives. But there is hope on the horizon, and we may soon be crossing the threshold of a new era where spaceflight is not only cheaper and more frequent, but much safer and more ambitious than at any time since Apollo.
I. The First Space Age: Failure is Not an Option
Only a decade removed from the megadeaths of WW2, the Cold War superpowers were predictably callous about the human cost of their technological programs: Members of the Armed Forces or even, sometimes, the general public were subjected to lethal medical experiments, atomic bomb tests, and other harmful studies - often without their consent. Most relevant to this discussion, the death toll of experimental aviation technology was legendary, as illustrated by the famous Tom Wolfe book (and later film adaptation) The Right Stuff: Finding out whether something worked or not consisted of putting a human pilot in it and having them fly it beyond its design limits, then either recover and land or bail out.
Test pilots would do this over, and over, and over, testing a multitude of different properties of the aircraft, with the engineers working on the technology changing things as they went along based on what the pilots reported. This is still basically how things work with experimental aircraft, but today the approach is far more subtle: Countless computer simulations and unmanned tests precede putting a pilot in the cockpit, and even then they methodically tease the limits over test programs extending over years. In the late 1950s and early '60s, however, it was considered easier and quicker to just put a man in the cockpit and tell him to push it as far as he could, and a smoldering pile of wreckage on the desert floor could be almost as informative as a living pilot's reports. With the resources of the entire Cold War US economy behind these efforts, experimental aircraft were cheap and lives were cheaper.
But that all changed when this same corps of elite test pilots became astronauts: The task was essentially the same, although in a totally unprecedented domain - to operate the first generation of manned spacecraft so that American technology could evolve to meet the challenge of beating the Soviet Union to the Moon. And because of the mythos surrounding their missions, and their status as national heroes and symbols, they were no longer expendable. Not even the legendarily callous Soviet government could get around it - they too discovered that the people riding the rockets were politically indispensable. Thus Failure Is Not An Option became the ideological credo of the First Space Age well before Gene Kranz first stated it, and has remained to this day.
When you think about it, it made little practical sense: The rockets the first astronauts and cosmonauts flew on were little more than improved ICBMs (which both countries were churning out like sausages), and the capsules they rode in were just airtight tanks with a few bells and whistles attached - very crude stuff. And moreover, every last system other than the pilot was expendable, and would be used only once. So there was no economic imperative to insist that no expense be spared to ensure success every time - quite the opposite. Either or both superpower programs could have advanced much farther, much more quickly if they had taken the same approach to spaceflight as military aviation. But they couldn't, because the eyes of the world were on them and every failure was a national failure - every loss of crew a national tragedy that cast doubt on the entire program and the competence of the government managing it.
But, ironically, Failure is Not an Option comes at a heavy price not just in money and time, but in lives and the mission for which those lives are sacrificed. To be as certain as possible of success before building up experience and learning hard practical lessons, both the US and Soviet programs had to design their entire launch architectures around enormously expensive and inefficient systems that require standing armies to manage and operate and years of lead time to launch anything. And what that means is you can't launch very often: Every single human space launch from 1961 to this very day has had to be managed like a mini-space program unto itself, involving an unimaginable level of complexity and potential failures.
Because Failure is Not an Option, there was simply no opportunity to build up experience in a timely fashion: Lessons that might have been learned in a year were instead only learned in ten years at orders of magnitude greater financial cost and the same if not higher ultimate number of deaths. And this is where the Apollo 1 fire in 1967 that killed astronauts Gus Grissom, Ed White, and Roger Chaffee provides an important example: All the billions of dollars and an entire decade of preparation that had gone into the program up to that point failed to account for a trivial matter of air mixture - something that could have destroyed a space mission while underway, which might easily have made it impossible to find out what happened or why.
The air mixture in the Apollo 1 capsule was not something new: All the Mercury and Gemini flights had used pure oxygen, and the full-up ground tests that preceded them had all used it at greater than ambient pressure, so the risk of fire had been there all along. But because these tests and flights were so infrequent, the program transitioned from 1 astronaut per mission to 2, then from 2 to 3, and thus when the probabilities finally added up to a fire, it was three astronauts rather than 2, 1, or none at all that were killed in order to learn this lesson. When you insist that Failure is Not an Option, the cost in lives of the failures that ultimately do happen is going to be higher, and the programs they die for will take longer, cost more, and have less to show for it.
Now, granted, this approach did get America to the Moon and bring every single one of its lunar explorers home alive, but the entire system had been developed on a totally unsustainable footing - every mission was tremendously expensive, time-consuming, and infrequent. And so, having sacrificed not only the three Apollo 1 astronauts to reach the Moon, but a number of others in training accidents, as well as all the ground crew, technicians, and engineers both at NASA and its contractors who had died for the program over the years, America simply abandoned the Moon three years after first landing on it and has not returned since. Because the failure of a single mission was not an option, the failure of the entire program was guaranteed, and the sacrifices made for it were simply discarded.
One of the multitude of ironies surrounding this fact is that the purposes for which the Air Force spent so many lives in its test programs were not even in the same ballpark of importance as Apollo, but largely thanks to those sacrifices, aviation - both military and commercial - advanced by leaps and bounds in the decades following WW2. We went from propeller-driven tin cans flying 300 mph at 5,000 feet with a dozen passengers to pressurized jumbo jets flying hundreds of people 500 mph at 35,000 feet in a single generation. But because the lives of the astronauts were so politically precious, and the public perception of even unmanned technical failures so damaging, Apollo was doomed to be abandoned in its moment of triumph and NASA would be left to wander aimlessly in low Earth orbit up to this very day.
In fact, we can say that Failure Is Not An Option killed Apollo not only through cost and lack of cumulative experience, but as a matter of direct necessity: Exploring the Moon with the technology developed under that philosophy was never going to be safe enough to satisfy it, so it was decided to just quit while we were ahead and count ourselves lucky that no one had died on the Moon or in transit. Got that? Because safety was too important for space missions to happen frequently, the only architecture that could be developed within the decade time limit was too unsafe to keep using, and thus our ambitions had to be scaled back rather than changing the philosophy that had sabotaged them in the first place. Such crazy thinking would be the hallmark of the next period in NASA history, the Shuttle program, and would play a central role in both of the Shuttle's catastrophic disasters.
"FINAO" in practice turns out to not so much be a statement of rational priorities, but more a state of denial that simply defers the inevitable human cost of exploration while sabotaging its progress. We've all met people who are so obsessed with neurotic fears of death that they forget to actually live; people who spend so much time on their appearance that they have nothing else to offer beyond initial attraction; cars that are so cheap, they end up costing more because of all the damn maintenance, etc. Well, that's kind of what happened to NASA, and why human spaceflight remains dangerous, expensive, and infrequent: By political necessity, they were (and remain) so focused on not failing that they ended up forgetting how to go anywhere or do anything, and yet catastrophes still happened - in fact, more of them, with a higher death toll, and on behalf of increasingly trivial missions.
I suppose we should be grateful that "FINAO" at least kept manned spaceflight on life support, since actually ending it altogether would be "failure" and thus not an option. But instead of outright euthanasia, we got an induced coma - "space exploration" was redefined to mean nothing more than "doing random busywork where the sky is black." No destination or long-term purpose needed. No technological or economic improvements need apply. And thus, the Space Shuttle was born: More expensive to fly than any other rocket, and statistically more dangerous to astronauts than any other system they had ever ridden to space, but it looked cool and carried lots of miscellaneous things to do so the public would forget nothing was actually being accomplished. At this point, NASA started pushing the idea that the purpose of space exploration was science, when before then it had clearly been the other way around.
This made perfect political sense, of course: Science as the purpose of the program could be done in low Earth orbit, and there would be no opportunity to apply what was learned in ambitious ways that might result in failure. The single boldest thing the Shuttle ever did was to fix the Hubble telescope in 1993 - i.e., correct a failure that had already occurred, because FINAO. Except that, oops, once again it was a failure caused by FINAO itself: All the expensive, time-consuming ground-based preparation for the Hubble trying to create a perfect product had instead yielded blurry optics requiring another expensive, time-consuming mission to correct.
In fact, they had no choice about doing it that way: NASA had to apply FINAO to Hubble and every other probe because FINAO had made launch so expensive that it would be uneconomical to just shotgun-blast the heavens with cheap probes and let statistics sort out what works and what doesn't. So this self-perpetuating philosophy of stasis born in 1961 has been screwing both the manned and unmanned sectors of space exploration ever since. And in the final decade of its run ending in 2011, what was NASA using the Shuttle for? To construct a space station - not as a launching point to go anywhere else; not as a dry-dock to construct interplanetary spaceships; nope - just as another, bigger place to do science.
Often the science being done on the station isn't even designed to enable space exploration! Just any old thing where a weightless environment can be useful. Gee, we could test out an artificial gravity centrifuge that could enable long-term manned missions to Mars, but instead we'll host a bunch of pharmaceutical experiments for Pfizer. See, because exploration is dangerous, and danger risks failure, exploration is not an option. Pure science, however, is relatively hard to fail - especially in little modular, self-contained experiments shipped up and down like cargo.
This is the "mission" that Shuttles Challenger and Columbia were serving when they were destroyed, that fourteen astronauts who had believed in the dream of exploration ended up giving their lives for - busywork invented to serve the FINAO philosophy. All because of yet another horrible, ridiculous irony: The more banal and mundane the activity becomes, the less politically catastrophic it is to kill people in service to it. Because of FINAO, if those brave people had died on Mars, that would have been the end of that entire program. But because they died serving the status quo, doing miscellaneous science or maintenance chores in orbit for a bureaucracy that lacked the political clout to give them the exploration missions they all dreamed of, there was never any significant danger that the program would be canceled. Ever since the end of Apollo, the message from Congress has been loud and clear: Kill as many people as you want, as long as no one can figure out what the sacrifice was for.
Thanks to this madness, the Space Shuttle program turned out to be statistically more dangerous than its predecessors, and in its final years needed so much time between launches to be made "safe" that the flight rate collapsed to 1 or 2 per year while costs ballooned to unprecedented levels. So thanks to the politically-driven obsession with safety, decades have been squandered making space ever more dangerous, more expensive, and less frequent, and now the United States doesn't even have an independent manned spaceflight capability. That's right - FINAO degenerated our capabilities so badly that it became safer to ride antiquated Russian rockets than our own technology, and they're not getting any safer either. Thank God someone has decided to stop this crap and get serious...
II. The Second Space Age: If at First You Don't Succeed, Fly, Fly Again!
If an automaker in 1920 had been forced to make cars as safe as they are today, what they built would have looked like a tank, had a top speed of 5 mph, a range of 20 miles, and cost more than the owner's house. I'm exaggerating, of course, but you get the idea: It would have defeated the purpose of having a car, and might even have ended up being more dangerous due to how few of them would be made, and how rarely those few would be operated - unforeseen, catastrophic problems could crop up that had already been dealt with in more prolific, "unsafe" cars.
To make real progress on safety you first have to build a thing to accomplish its purpose: In the case of a 1920s car, transport people more quickly and over longer distances than a horse. Believe it or not, safety is a secondary consideration to accomplishing that mission - otherwise you wouldn't buy a car in the first place, because it's safer to just not travel. However, once you are able to build something that does the job with some reasonable (not absolute maximum) safety margin, it may be affordable and practical enough for a significant number of people to use them. Their experiences quickly add up to a massive amount of concurrent testing data that enables rapid, continuous advancement of every aspect of the technology, including safety. As the quality, affordability, and efficiency of the product improves, so the number of people using it will increase, yielding an even larger set of data, and so on.
After ninety years of that, and with regulators pushing manufacturers to adopt proven safety measures, you end up with cars as safe as they are today: Airbags, three-point seatbelts, crumple zones, ABS brakes, traction control, etc. etc. None of which would or could have been developed by an automaker who had been forced from the beginning to make cars as safe as possible, rather than as safe as practical. In a world where doing so had been the law, all the automotive technology of the last century would be concentrated in one or two manufacturers of very expensive products: It would all have gone into making these slow, massive, short-range tank cars work incrementally better and look incrementally better, all the roads would be designed for them, and anyone who came up with radical new ways to be just as safe while having far superior performance would have to single-handedly invent an entirely new industry because every automotive supply chain would be geared toward the lumbering hulks.
That's what happened with spaceflight: The entire space industrial edifice is designed around the hulking, hyper-expensive, uber-complex infrastructure built in the 1960s to meet the FINAO imperative, and there is simply no room in it for anything that doesn't also abide by that imperative. And that's why, rather than improving, both safety and cost have been getting worse since then in the systems designed around it. Enter one Elon Musk, founder, CEO, and CTO of SpaceX (and CEO of Tesla Motors). Musk has not only done exactly as mentioned - single-handedly inventing his own supply chains to circumvent a sclerotic industry - but done it twice, in the case of both space rockets and automobiles. His companies are no mere assemblers of other people's components: They are vertically integrated, meaning they not only make almost every component, but in a number of cases they make the tools that make those components.
As a result, if his teams want to change something in a design, they don't have to go through some ruinous process of trying to communicate and coordinate with disparate subcontractors and suppliers all over the world: They can just get up and walk to the relevant desks and have a conversation, or else plan to have a meeting an hour later - simple, quick, productive. Innovation incarnate. This is how SpaceX's current rocket, the Falcon 9, has already reduced launch costs from the rest of the industry by 2/3 while having a thus-far perfect success rate at 4 out of 4 (knock on wood). And that's just the beginning - they will soon be transitioning to a vastly more powerful and more efficient engine and larger fuel tank, and are also pursuing stage-reusability that would decimate costs and radically increase safety if successful.
And how is this possible? Well, because Musk spent the first six years of SpaceX history failing to launch rockets! He littered the equatorial Pacific with multimillion-dollar wreckage not once, not twice, but three times in a row, each time demoralizing the hell out of his employees, agonizing over what to change, and drifting ever closer to personal bankruptcy with each failure. But every time, failure was most definitely an option. And then...it worked. And whoa-ho did it ever work - it worked so well they were able to put 9 of them together to create the Falcon 9, and that has worked every time they've launched it. You may have heard they've also built what will become a manned spacecraft in a couple of years, the Dragon, and have flown it autonomously to the International Space Station twice, replicating key technical achievements of the Gemini program (e.g., docking in orbit) at a tiny fraction of the cost.
You may also have heard about a hiccup in the most recent Falcon 9 / Dragon cargo flight where the rocket lost an engine mid-flight: F9 was specifically designed to tolerate such a loss - because...you guessed it...failure is an option - and thus the mission still reached orbit and delivered the Dragon. Failure of components is an option in order to make mission success more likely, and mission failure (like the three original failed launches) is an option in order to make the success of the overall program more likely - a program that has as its ultimate goal the large-scale human colonization of Mars within two decades. Thanks to all this "failing," both at SpaceX and Tesla, Elon Musk is now a multi-billionaire (his net worth before had only been 9 figures), and is on track to rescue NASA manned spaceflight from its politically-induced death spiral.
Now, failures of unmanned rockets and satellites are one thing, but I don't want to sound like I'm saying that SpaceX would be lax in safety when it comes to human flights: Quite the opposite. What they've done is initiate as a "virtuous cycle" beginning with the unmanned launches where the rockets and spacecraft that will ultimately launch people are made ever-safer, ever-more reliable, and increasingly affordable, precisely because they're willing to accept failure as a cost of doing business. By the time people are using these rockets and spacecraft, they will have flown so many times, and ironed out so many issues, that it's likely the very first flight of the manned Dragon will be the safest human space launch in history. And as a result of both that safety and low cost, a lot more of these flights will happen per year than has ever been possible before.
NASA itself may never fully escape from the FINAO trap, since it is by nature bound to politics, but SpaceX doesn't need them to in order to deliver these advances: As a private contractor, they will also be launching commercial manned Dragon flights for rich private customers, corporations, universities, and any number of small countries around the world. And these flights won't necessarily be to the space station - so orbital flights can happen that don't need to be bounded by the ISS schedules and crew rotations, meaning they can fly at their convenience. This is especially true if SpaceX proceeds with its plans to build its own spaceport, so they wouldn't even be limited by the schedules of public-sector launch ranges like Cape Canaveral and Vandenberg AFB. In other words, they could launch as often as demand and their own ability to supply could handle.
More launches means more experience, more opportunities to deal with the technology in practice, and thus way more data to work with in designing improvements and subsequent generations of spacecraft - just like in the automotive example. Moreover, while obviously they would want some reasonable assurances, private customers looking for a space adventure or a research flight might not be as nitpicky about exactly how the desired safety margins are achieved: As long as they aren't endangering third parties, they can make their own cost/benefit judgments and assume greater risk if they deem it a better proposition. And the beauty of it is that NASA would benefit from all of it even if they themselves remain bound by law or bureaucracy to still insist on FINAO. The safety offered by Falcon 9 / Dragon would just keep improving thanks to the frequency of flights made possible by private customers.
There is, however, a down side - although no more than anything we do here on Earth: Even though the statistical chance of a deadly accident would go down, the fact that flights are happening a lot more often and taking a radically increased number of people into space means that sooner or later the absolute number of spaceflight deaths in a given period of time would keep increasing for generations. If Dragon is made many times safer than Shuttle, but flies many times more frequently, you end up with more casualties in terms of absolute number even though any given astronaut or passenger is a lot safer. This was a conundrum also encountered in aviation history, when planes started flying dozens of passengers rather than just one or two stunt pilots, and it's something we'll have to get used to seeing.
However, even that will eventually pass, as the lessons learned from high-volume spaceflight catch up to the growth in number of flights. At some point Earth orbit is bound to become as safe and routine as air travel is today, although that's generations away. Once that happens, the "zone of danger" would retreat out to "destination flights" beyond Earth - the Moon, Mars, Near Earth Asteroids - where the greater duration, impossibility of getting back to Earth quickly in an emergency, and the radiation hazards become the new domains where hard lessons will have to be learned at some human cost.
The bottom line is this: Given all the people who die pointlessly around the world through accidents, murders, and diseases - and the fact that we all die eventually, most of us without having accomplished much - there is no number of deaths for something as profound as human expansion into space that I find morally problematic. Astronauts are intelligent, courageous people risking their lives for a cause, and I can't imagine any level of informed, voluntary risk being too high if it substantively contributes to that cause. What I cannot stand for, however, is the waste of that courage and sacrifice by a government that won't let astronauts expand humanity's reach, but instead keeps risking them on missions to do piddling science experiments with no direct relevance to enabling space exploration.
No one should die for a laboratory, and certainly not for mere pork-barrel spending. When astronauts were children dreaming of space, they did not dream of being lab technicians in a so-called "outpost" a mere 250 miles from the ground, two generations after humans were roving the surface of the Moon. They dreamed of being explorers, and at the risk of being presumptuous, I can say with confidence that the opportunity to be explorers is what they signed up for when they joined NASA. What NASA has done with the sacrifices of the Apollo 1, Challenger, and Columbia crews is a disgrace and a desecration, as well as an insult to everyone who has ever been inspired by its past achievements and deceived by its rhetoric of bold exploration.
But I don't blame NASA itself - it's mostly Congress that's responsible for its many disgraces over the years - and I think the agency to some extent realized it was wasting away after Columbia. That's why its official plans include the construction of the Orion space capsule designed for missions to lunar orbit and asteroid rendezvous and the Space Launch System (SLS) rocket to propel it to these far destinations beginning in 2021. Even if these projects are canceled, more advanced versions of Dragon are planned that would be capable of taking up the reins, so it's unknown at this point how things will shake out. Maybe NASA is finally getting its act together...
Still, I don't know how much if any faith to place in Orion/SLS plans, or in the eventual safety of these systems if they are completed and used as advertised, but at very least if deaths occur on real exploration missions, there will be no question that those who died were paving a road to a brighter and more expansive future for humanity - not being sent in circles by a cowardly bureaucracy and a corrupt, nihilistic Congress.