When my father reached his 80s, I realized I’d never asked him much about his career as an engineer with NASA. I set out to learn all I could about the day-to-day worries, stumbles and triumphs he and his Apollo-era colleagues experienced. My first conversations with him set off years of research, a chain of interviews and eventually a book, The Apollo Chronicles: Engineering America’s First Moon Missions.

He hauled out a box of mementos. From the pins, patches and plaques, he pulled out one of his 1960s research articles. “Whoa,” I whispered. It was crowded with matrices and equations. At a time when a voyage away from the Earth was still just a dream, this paper detailed the mathematical transfer of a spacecraft from an Earth-centered coordinate origin to that of another world.

When John F. Kennedy announced his moon challenge to the nation in 1961, thousands of minds set fervently to work. Incredulous, they gazed at legions of imposing problems: unsteady, exploding rockets; a largely mysterious lunar surface; and trajectory calculations full of best guesses. With less than nine years to complete a lunar landing, the engineers didn’t have a script. As my father’s colleague Marlowe Cassetti put it, “I can think of a toddler taking a first couple of steps and being unsure, and then you say, ‘In a couple of years, he’s going to run a marathon.’”

I find NASA’s scramble of 1962 especially inspiring. Waves of employees like Cassetti moved from Langley, Virginia, to Houston. Leaving scenic Virginia, some engineers found their new setting otherworldly enough, with scrublands flat as graph paper. “It nearly moved me to tears,” said Cassetti. “Gloomy, dreary, ugly,” recalled another transplanted engineer, Hal Beck. Both had been born and raised on the East Coast. And like so many young people at the headwaters of NASA, they had both opted for engineering after childhood obsessions with model airplanes. Now, in their 20s, they were planning space missions.

The new NASA center, rising from a muddy cattle pasture, wouldn’t be ready until 1964. In the meantime, the agency rented a smattering of buildings and suites around the greater Houston area. In some cases, apartment bedrooms and kitchens became makeshift offices. And without their own armoire-sized computers chugging away, the engineers had to borrow processing time wherever they could find it. In 1962, NASA leaned on an IBM 7090 computer at the University of Houston. Advanced for its time, with transistors instead of tubes, it still required instructions fed to it one punch card at a time, and it returned results the next morning.

One of Cassetti’s jobs involved computing the exact time an orbiting astronaut should start his capsule’s fiery reentry through the atmosphere. A perfect splashdown needed to strike the Earth in daylight, near (but not too near) waiting recovery ships, and without cratering into some landmass or another. Cassetti recalls carrying his ordered stack of punch cards—his computer program—through driving summer rainstorms, with runoff so strong it lifted manhole covers from the streets.

Even if he kept his cards dry, he often met with frustration when he returned the next day. “There was a high failure rate,” he said. “You’d punch one number wrong into a card, and you got a bunch of wasted paper … just a bunch of octal numbers that didn’t mean anything.” If he threaded the needle—punching his cards correctly and keeping them in order—he still suffered at the whim of the computer facility’s operators. Local staff ran the programs overnight, but they didn’t share the every-day-counts drive of the engineers. “Occasionally I would tell our guys, we got to have this thing run by tomorrow. Wrap the cards in a five-dollar bill if that helps.”

Meanwhile, Beck pursued one of the first step-by-step computations to the moon itself. He describes their calculation as a stumble through the dark. Ever-shifting gravity from two spheres, dynamic fuel levels and the changing weight of a spacecraft made for a scenario so complicated that there was no single equation to capture it all. Beck’s team sweated a laborious, step-by-step path to the moon, computing a range of parameters for every time point. “When you computed a lunar trajectory,” Beck said. “You’d have a stack of paper like a foot thick.” In 1962 they were not at all sure this approach was going to work on a computer, never mind for a real ship crammed with three astronauts.

Like Cassetti, Beck desperately needed his card stacks run each night at the university computer center. And he looks back fondly on his secret weapon. “It was a sweet car, man.” When dropping off his requests for the evening, Beck handed over a key to the staff along with his punch cards. Then he took a taxi home. As the staff decided whose jobs to run, they could stroll to the parking lot and find, in the Austin Healey’s trunk, a cooler full of beer cans on ice. And Beck’s team routinely had the thickest stack of results waiting each morning.

In the end, the toddler of NASA stayed on its wobbly feet and eventually started running. The engineers survived “core dumps” and rainstorms to nail capsule reentry times. They not only found the numeric razor’s edge to the moon, but also choreographed an awkward orbital dance between a fragile little ship and the moon’s lumpy gravity. Eventually, they timed a landing with optimal shadow lengths for the photos, films and first steps of astronauts.

When Apollo 11 settled to the moon, 50 years ago, I was larval, just a few months old. The moon landing has always been old news for the majority of my generation. Visiting the Johnson Space Center as a boy, I thought the retired Saturn V rocket, spread stage by stage on the ground, resembled the bones of a great dinosaur. Returning now to his mementos, my father pauses over the yellowed pages of mathematics. “I can’t believe I used to do things like this,” he says, shaking his head.