SpaceX and the science of failure

In the fields of science and technology in general, failure can be the key to success.

When SpaceX’s Starship exploded not long after launch last month, it was generally seen as a failure. But for SpaceX, and for science and technology in general, failure can be the key to success.

Move fast and break things

SpaceX founder Elon Musk was sanguine after his big new rocket suffered a “rapid unscheduled disassembly.” He embraces the tech sector’s “move fast and break things” ethos, first espoused by Mark Zuckerberg. He knows that the engineering data gathered from the short flight will help track down problems and improve the outcome the next time Starship launches.

This type of rapid innovation cycle has worked for SpaceX. The company’s first rocket was the Falcon 1. Its first three launches were failures, followed by two successes — a fairly low success rate. Then Falcon 9 became the workhorse rocket for a decade, with just two failures in 232 launches, for a phenomenal success rate of 99 percent.

Government space agencies move more methodically than SpaceX. Funded by the taxpayers, they can’t afford too many high-profile explosions. But even they learn from failure. In the first few decades of Mars exploration, half of all the missions failed, which prompted NASA scientist John Cassini to speculate that a galactic “ghoul” was dining on spacecraft aimed at Mars. But since 2000, the success rate has been nearly 80 percent. We’ve become accustomed to nimble rovers, sky cranes and helicopters operating seamlessly on the red planet.

The path of science

The public perception of science as a constant upward trajectory toward greater knowledge is wrong. Science is full of setbacks, twists and turns, and dead ends. Nearly everything that happens in the lab will never make it into print.

Then again, even mistakes can lead to discoveries. In 1884, a dentist was playing with nitrous oxide, commonly called laughing gas, at a party when he fell and gashed his leg. This led to its use to relieve pain in surgery. Scottish chemist Alexander Fleming’s sloppy lab habits led to the discovery of penicillin. In 1934, a bored chemist at DuPont was playing with a liquid polymer in a beaker when he realized he could draw it out into a thin solid strand. This was the discovery of nylon.

But scientists don’t generally blunder into knowledge. They purposefully form hypotheses and test them with the best data they can gather. The process of converging on the truth involves trial and error. As Nobel prize-winning physicist Richard Feynman said in a speech about how we learn about the laws of nature: “We’re trying to prove ourselves wrong as quickly as possible. Because only in that way do we find progress.”

The importance of null results

In science, any experiment that “fails” can be a success if it is well-designed. Astronomers and physicists celebrate new discoveries such as gravity waves and the Higgs boson. But much progress is made by painstaking observations that fail to discover anything. Null results can rule out theories and guide the design of future experiments.

In 1887, the Michelson-Morley physics experiment failed to detect ether, the invisible medium through which light was supposed to travel. This null result showed that the speed of light was a universal constant, and it impelled Albert Einstein to his special theory of relativity and a revolution in physics.

The problem is that null results don’t make you famous. Consequently, scientists don’t write up their failed experiments or attempts to replicate other scientists’ results, and journals tend not to publish them. It’s hard to get grants and attract students unless you discover something. The result is duplicated effort and bias in favor of marginal results as scientists chase weak signals that barely poke up above the noise.

Failure is personal

Nobody likes to fail. But science is very competitive; most grant applications fail, and most research papers are rejected. This rejection culture has some serious downsides. Mental health problems in academia are at an all-time high, at twice the rate of the general population. Attrition is high — half the young people working toward academic science careers will drop out after five years.

Scientists polish their CVs and fill them with accomplishments. However, there’s a small but refreshing trend to shine a light on failure. In 2016, Princeton professor Johannes Haushofer put his “CV of Failures” online. It listed every exam he had failed, every job application he’d been rejected for, every project that had failed, and every paper that had been rejected for publication.

To his surprise, it went viral.

Other scientists have now covered the walls of their offices with rejection letters, and that has gone viral as well. Failure is finally becoming a badge of honor.

Chris Impey is a professor of astronomy at the University of Arizona. He is the author of hundreds of research papers on observational cosmology and education, and he has written popular books on black holes, the future of space travel, teaching cosmology to Buddhist monks, exoplanets, how the universe began, and how the universe will end. His massive open online courses have enrolled over 400,000 people.