The most expensive hyphen in history
Bugs, bugs bugs
By Charles Fishman4 minute Read
This is the 18th in an exclusive series of 50 articles, one published each day until July 20, exploring the 50th anniversary of the first-ever Moon landing. You can check out 50 Days to the Moon here every day.
In the dark on Sunday morning, July 22, 1962, NASA launched the first-ever U.S. interplanetary space probe: Mariner 1, headed for Venus, Earth’s neighbor closer to the Sun.
Mariner 1 was launched atop a 103-foot-tall Atlas-Agena rocket at 5:21 a.m. EDT. For 3 minutes and 32 seconds, it rose perfectly, accelerating to the edge of space, nearly 100 miles up.
But at that moment, Mariner 1 started to veer in odd, unplanned ways, first aiming northwest, then pointing nose down. The rocket was out of control and headed for the shipping lanes of the North Atlantic. Four minutes and 50 seconds into flight, a range safety officer at Cape Canaveral—in an effort to prevent the rocket from hitting people or land—flipped two switches, and explosives in the Atlas blew the rocket apart in a spectacular cascade of fireworks visible back in Florida.
The Mariner 1 probe itself was blown free of the debris, and its radio transponder continued to ping flight control for another 67 seconds, until it hit the Atlantic Ocean.
This was the third failed probe in 1962 alone; NASA had also launched two failed probes to the Moon. But the disappointment was softened by the fact that a second, identical Mariner spacecraft (along with an identical Atlas-Agena rocket) were already in hangers at the Cape, standing by. Mariner 2 was launched successfully a month later and reached Venus on December 14, 1962, where it discovered that the temperature was 797º F and that the planet rotated in the opposite direction of Earth and Mars. The Sun on Venus rises in the West.
It was possible to launch Mariner 1’s twin just 36 days after the disaster because it took scientists at NASA’s Jet Propulsion Laboratory only five days to figure out what had gone wrong. In handwritten computer coding instructions, in dozens and dozens of lines of flight guidance equations, a single letter had been written incorrectly, probably forgetfully.
In a critical spot, the equations contained an “R” symbol (for “radius”). The “R” was supposed to have a bar over it, indicating a “smoothing” function; the line told the guidance computer to average the data it was receiving and to ignore what was likely to be spurious data. But as written and then coded onto punch cards and into the guidance computer, the “R” didn’t have a bar over it. The “R-bar” became simply “R.”
As it happened, on launch, Mariner 1 briefly lost guidance-lock with the ground, which was not uncommon. The rocket was supposed to follow its course until guidance-lock was re-achieved, unless it received instructions from the ground computer. But without the R-bar, the ground computer got confused about Mariner 1’s performance, thought it was off course, and started sending signals to the rocket to “correct” its course, instructions that weren’t necessary—and weren’t correct.
Therefore “phantom erratic behavior” became “actual erratic behavior,” as one analyst wrote. In the minute or so that controllers waited, the rocket and the guidance computer on the ground were never able to get themselves sorted out, because the “averaging” function that would have kept the rocket on course wasn’t programmed into the computer. And so the range safety officer did his job.
A single handwritten line, the length of a hyphen, doomed the most elaborate spaceship the U.S. had until then designed, along with its launch rocket. Or rather, the absence of that bar doomed it. The error cost $18.5 million ($156 million today).
In the popular press, for simplicity, the missing bar became a hyphen. The New York Times front-page headline was “For Want of a Hyphen Venus Rocket Is Lost.” The Los Angeles Times headline: “‘Hyphen’ Blows Up Rocket.” The science fiction writer Arthur C. Clarke, in his 1968 book The Promise of Space, called it “the most expensive hyphen in history.”
For NASA’s computer programmers, it was a lesson in care, caution, and testing that ended up steeped into their bones. During 11 Apollo missions, more than 100 days total of spaceflight, the Apollo flight computers performed without a single fault.
But what happened to Mariner 1 was, in fact, an arresting vulnerability of the new Space Age. A single missing bolt in a B-52 nuclear bomber wasn’t going to bring down the plane, but a single inattentive moment in computer programming—of the sort anyone can imagine having—could have a cascade of consequences.
George Mueller was NASA’s associate administrator for manned spaceflight from 1963 to 1969, the most critical period for Apollo’s development. Just before that, Mueller had been an executive at Space Technology Laboratories, which had responsibility for writing the guidance equations for Mariner 1, including the equation with the missing bar.
During his years at NASA, Mueller kept a reminder of the importance of even the smallest elements of spaceflight on the wall behind his desk: a framed image of a hyphen.