"Adrian Cole," the younger man said, offering his hand. "Systems dynamics."
"Isaac."
"I read your 2017 report."
"Sorry to hear that."
Adrian blinked.
Isaac turned toward the monitors.
"Show me the last four hundred orbital cycles."
The first hour went badly for Adrian's assumptions.
He began by presenting the standard decomposition of thermal response, actuator load, and pressure instability. Isaac let him finish. Then he pointed to a narrow band of data Adrian had classified as low-confidence noise.
"Why'd you throw that out?"
"Signal-to-noise ratio."
"That's not an answer."
"It's below the confidence threshold."
"Why?"
Adrian hesitated.
"Because the sensor package isn't calibrated for useful interpretation at that amplitude."
"Good. Now forget the amplitude. What about the timing?"
Adrian looked again.
The low-confidence feature repeated.
"Thirty-one minutes."
"And?"
"One third of the orbital period."
"And?"
Adrian's eyes narrowed.
"It's phase-locked."
Isaac nodded.
That was all.
No speech about being overlooked. No triumph.
Adrian spent the next forty minutes rebuilding his analysis around timing rather than amplitude.
At lunch, Eleanor found him sitting on an overturned bucket with a sandwich in one hand and a notebook balanced on his knee.
"How's it going?"
Adrian looked toward Isaac, who was beneath a customer's tractor near the open bay door while a simulation ran.
"I think I've been weighting expensive sensors more heavily than cheap questions."
Eleanor smiled. "That sounds healthy."
"He doesn't explain unless I answer first."
"Also healthy."
"It's irritating."
"Especially healthy."
By the second afternoon, the team had a complete working theory.
The valve assembly was not simply degrading from repeated temperature swings. The orbital thermal transition altered the mechanical stiffness of the housing enough to shift its natural response into alignment with a recurring actuator excitation. For a narrow part of each orbit, the system effectively shook itself at the worst possible frequency.
The effect was small per cycle.
That was what made it dangerous.
Catastrophic failures attracted attention. Slow failures adapted to organizational habits. Each day's telemetry looked only slightly worse than the day before. Individual limits remained inside specification until the accumulated damage became difficult to reverse.
"It's a debt system," Isaac said.
Adrian looked over. "Mechanical fatigue?"
"Institutional thinking. The hardware borrows margin every orbit. Nobody cares because the payment is tiny. Then one day the account's empty."
Adrian wrote the sentence down.
"Don't put that in the report."
"Why not?"
"Because they'll call it metaphor and ignore the numbers."
The first proposed remediation came from Nexora's current engineering team.
They recommended modifying the regulator's command cycle to reduce actuation frequency during the most damaging thermal transition. On paper, the approach reduced vibration amplitude quickly and required minimal change to existing command logic.
Isaac read the package twice.
Then he asked Adrian for the mission propellant budget.