Space to Fail

L ast week, Blue Origin’s New Glenn rocket, capable of delivering payloads to Earth orbit and beyond, exploded spectacularly on its launch pad in Florida. My phone immediately lit up, as fellow space nerds lamented the fact that we were unlikely to meet the target of putting an American back on the surface of the Moon by the end of Trump’s second term. Just a few days earlier, NASA had announced additional awards to Blue Origin (who has been working on a large robotic lunar lander for many years on its own dime) to deliver the first elements of the Moon Base program to the lunar surface.

There is no doubt that this is a major setback for the company, but already, both NASA Administrator Jared Isaacman and Blue Origin CEO Dave Limp have expressed confidence in a quicker-than-expected recovery, with launches resuming as early as late this year.

The frenzied response to the failure has once again exposed the simmering rifts in how we, as a nation, approach space exploration. The most obvious contrast is between the NASA of the late 20th century — featuring the Space Shuttle and International Space Station programs — and the emerging NASA of the 21st century, which has seen the rise of partnerships with SpaceX, Blue Origin, and other “new space” companies.

As of today, SpaceX dominates the launch vehicle market. Last year alone, SpaceX and its reusable Falcon 9 technology outpaced all Chinese launches by a rate of almost 2-to-1. This is remarkable, given that as recently as 2008, it was not clear that SpaceX would even survive as a company given its early failures with Falcon 1.

Back then, as an aerospace engineer at NASA’s Johnson Space Center, it was easy for me to poke fun at Elon Musk and the upstart SpaceX. The common wisdom was that SpaceX was swimming in deep waters that it did not understand and would fail just like other space companies did before it — thus proving the saying, “if you want to become a millionaire in the space business, better start as a billionaire.”

What I did not realize then was that taking a risk and ultimately experiencing failure was a necessary feature, not a bug, of what has become the most-flown American rocket in history. Why is SpaceX’s Falcon 9 rocket so ubiquitous today? While it’s true SpaceX was swimming in deep waters, the company’s strategy was to learn faster than anybody else in the industry — hearkening back to NASA of the Apollo days.

In 1962, nobody thought sending humans to the Moon — never mind returning them safely before the end of the decade — was possible.  And now, the Falcon 9 rocket has achieved a capability no one at NASA thought was possible: rapid reusability. This has allowed SpaceX to save billions of dollars in manufacturing costs and is now the most reliable rocket on the planet.

How did SpaceX accomplish something that today’s NASA thought was not worth the risk? The answer is in how teams relate to and respond to failure. Over time, people and organizations become risk-averse. NASA is no exception. In the early days of the agency, with the threat of a highly capable Soviet Union pouring millions into its own space ambitions, NASA had many spectacular failures. Fortunately for NASA, its earliest failures did not result in the loss of life.

That all changed on January 27, 1967. On that day, a fire broke out in the pure oxygen cabin of the Apollo 1 spacecraft. Without the ability to do a rapid egress from the burning capsule, all three crew members — Gus Grissom, Ed White, and Roger Chaffee — perished. The Apollo 1 fire could have ended the program before it could get started.

But the Cold War was not an ordinary time, so the nation pushed through the tragedy and successfully put a crew into orbit with the Apollo 7 mission. Good lessons were learned from that event, including new methods to mitigate and reduce risk to human lives. However, NASA understood that it could not completely eliminate risk — spaceflight is inherently risky — and continued to rapidly innovate in order to put humans on the Moon before the decade was out.

Having successfully landed Neil Armstrong and Buzz Aldrin on the Moon with Apollo 11 in 1969, and then the success of Apollo 12, it seemed that Apollo 13 should be a walk in the park when it launched in 1970. But then an explosion catastrophically damaged the Apollo service module three hours into the mission, and the chances of returning the brave crew looked grim.

During that mission, the heroics of NASA personnel such as Gene Kranz brought the crew back to Earth safely. During that mission, Kranz coined the phrase “failure is not an option.” He even published a book by the same name. This kind of thinking spread throughout NASA’s culture. It was a luxury the United States could now afford, given that we had already beaten the Soviets to the Moon. It was also the beginning of the end of NASA’s risk-taking culture that prioritized speed and innovation.

While “failure is not an option” is a good goal for a team conducting real-time operations, that motto is the opposite of what a company needs during new system development. The key to rapid innovation is to test and, yes, fail. In Silicon Valley, they call this “rapid do-learn loops,” or “fail-fast-forward.” The focus is on maximizing and accelerating learning. That is how America has become the world leader in technology.

Today, American entrepreneurs understand this, and so do our venture capitalists. But this is not new — the NASA of the 1960s knew this, as did Kelly Johnson’s Skunk Works at Lockheed Martin, as well as Bill Boeing and Donald Douglas’s aircraft companies in the 1920s and 1930s.

The logical outgrowth of the “failure is not an option” culture is to rely on “known” legacy hardware, layered on top of excessive requirements, and triple-and-quadruple-checking everything. In contrast to SpaceX, the late 20th and early 21st centuries have seen traditional companies embrace this approach, as it allowed them to continue to build and sell systems they had already developed and permitted NASA to build redundancy into the integrated systems — which comes with higher costs and longer development timelines.

The result? Systems that are expensive and slow and thus infrequently launched and, counterintuitively, that have directly affected NASA’s ability to ensure future reliability and repeatability. Expensive systems that don’t fly often are unsafe.

Today, we are at a new crossroads. We can choose to accept the risk of failure by encouraging adoption of new, innovative approaches to space exploration, which are actually old approaches proven many decades ago. Alternatively, we can choose to continue pursuing traditional approaches that we understand and are comfortable with, but we know are expensive and time-consuming — thus preventing the rapid emergence of cheaper and more frequent access to space.