Read Chapter 20: THE UNMANNED FAILURE

The final piece arrived in 1971.

Luc was reviewing archived anomaly reports as part of a reliability study when he found the unmanned drop test Reed had mentioned years earlier.

The official cause remained connector damage during handling.

But the original telemetry had recently been transferred to a central archive under the new raw-data retention policy.

Luc requested the tape.

The connector channel remained continuous through the event.

The DCP output, however, dropped for 4.6 milliseconds.

He sat alone in a playback room listening to the tape transport spin.

There it was.

The earlier failure.

Before Galveston. Before thermal vacuum. Before Delaney’s notebook entry.

Asterion had known the failure mode had appeared in a dynamic vehicle test.

Luc traced the report history and found that Frank Delaney had written the first anomaly draft. His original cause assessment identified “probable intermittent relay contact under vibration.”

A later revision changed the cause to “connector disturbance induced during handling.”

The approving signature belonged to Nathan Vale.

Luc called Hale, now back in Washington.

“The trial is over,” she reminded him after hearing the details.

“This would have strengthened it.”

“Yes.”

“Can the record be corrected?”

“Technically, yes.”

“Then do it.”

“Luc, nobody is flying that configuration. Vale has been convicted. Asterion has restructured.”

“The record is false.”

Hale sighed.

“You really have learned nothing about choosing battles.”

“I learned exactly one thing.”

“What?”

“Bad records outlive the hardware.”

She laughed softly. “Send me the package.”

NASA amended the anomaly report six months later.

The change attracted no press.

No hearing.

No trial.

One sentence in an archive now said what had happened.

Luc considered it one of the most important victories of his career.

The archived drop-test telemetry revealed more than the relay failure.

It showed that Delaney’s initial diagnosis had been technically sound even without the supplier report. He had compared connector continuity, regulator input and output, and vibration timing. The connector remained stable. Only the conditioned output disappeared.

Luc reconstructed the analysis and realized Frank had effectively solved the failure two years before Luc arrived in Houston.

The fact humbled him.

Professional hierarchies had hidden not only risk but credit. Delaney’s finding had been rewritten by managers because his conclusion was inconvenient, then rediscovered later by engineers with more authority.

Luc added a new requirement to NASA review practice: initial anomaly author and dissenting technical assessments should remain attached to the closure record, even when final cause differed.

“Why preserve wrong guesses?” a manager asked.

“Because sometimes they are not wrong.”

“And if they are?”

“They show what evidence was available at the time.”

This mattered for future investigators. A polished final report often created the illusion that the accepted explanation had always been obvious. Preserving disagreement showed the path, including roads not taken.

History, Luc had learned, became dangerous when edited to look inevitable.