Victoria turned toward the other engineers.
“Does anyone have a better answer?”
Nobody spoke.
“Does anyone have a theory that does not begin with the words 'we believe'?”
One of the younger engineers looked down.
Victoria pointed at him.
“Dr. Patel. You ran the last calibration. What did you discover?”
Dr. Patel cleared his throat.
“The engine operates normally at reduced load. When we increase the load, the temperature begins rising. The sensor data indicates that the cooling regulator is responding, but the temperature does not stabilize.”
“So the regulator is defective.”
“Not necessarily. It may be responding correctly to an inaccurate signal.”
Victoria stared at him.
“Then why haven't you corrected the signal?”
“We need to verify the sensor's frequency response under full operating load.”
“Have you done that?”
“Not yet.”
“Why not?”
“Because the previous calibration procedure—”
Victoria raised her hand.
“Stop.”
Dr. Patel fell silent.
“I do not need another explanation of what you have not done. I need to know when you will do it.”
“Tomorrow, if we can obtain the necessary test equipment.”
“Today.”
“The equipment is being used in another laboratory.”
“Then move it.”
Dr. Patel hesitated.
Victoria's voice sharpened.
“Is there a reason you cannot follow a direct instruction?”
“No, ma'am.”
“Good. Then I expect results before the end of the day.”
The meeting ended with a new deadline, two additional tests, and an instruction that no engineer could leave the building until the team had identified a plausible corrective action.
As the engineers filed out, Daniel remained behind.
“Victoria, we need to discuss the schedule.”
She did not look up from her tablet.
“What about it?”
“The current deadline is forcing the team to repeat tests before we have finished analyzing the previous results. We are introducing additional variables without understanding what the earlier tests have shown.”
Victoria continued reading.
“Your team has had six weeks.”
“Yes.”
“And the engine is still broken.”
“Yes.”
“Then I suggest you use the time you have left more effectively.”
Daniel took a breath.
“I am asking for one additional day to complete a controlled investigation.”
Victoria finally looked up.
“You are asking me for more time after telling me for six weeks that the solution is almost ready.”
“I am asking for enough time to verify the diagnosis.”
“And if you are wrong?”
“Then we will have eliminated another possibility.”
Victoria stared at him.
“Do you understand what is at stake?”
“Very clearly.”
“Fifty million dollars.”
“Yes.”
“Three major contracts.”
“Yes.”
“My credibility with the board.”
Daniel paused.
“That is also at stake. But if we rush the investigation, we may lose more than the contracts.”
For a moment, Victoria looked almost offended by the suggestion that her own pressure might be contributing to the problem.
Then she stood.
“You have until tomorrow morning.”
Daniel nodded.
“Thank you.”
“Do not thank me yet.”
He left the room.
Victoria remained beside the machine, staring at the exposed components.
She knew the pressure was affecting the team.
She also knew that reducing it would look like weakness.
Investors expected certainty. The board expected leadership. Employees expected her to solve the crisis.
If she admitted that she did not know what was wrong, she would have to face the possibility that her confidence had been premature.
She would have to tell the board that the schedule she had promised might be impossible.
She would have to explain that a project she had described as nearly complete remained fundamentally unreliable.
Victoria had built her career by avoiding that kind of admission.
She would not begin now.
At least, that was what she told herself.
Across the building, on the lower service level, Jamal Washington was replacing a damaged floor panel outside the materials laboratory.
Nobody had told him that the engine was failing again.
Nobody needed to.
The entire building seemed to carry the tension.
Engineers walked through the corridors with tired faces. Supervisors spoke in clipped sentences. The elevators were crowded at unusual hours because members of the development team were sleeping in offices and using the showers near the fitness center.
Jamal had worked in buildings long enough to recognize the signs of a crisis.
He also knew that the people closest to a problem were not always the people most likely to understand it.
His official job description included cleaning floors, emptying bins, maintaining supply rooms, and reporting obvious equipment hazards.
The position had originally been created as part of a contractor arrangement. Jamal's employment agency had listed him as a technical maintenance assistant because he possessed a basic engineering qualification and had completed several workplace safety courses.
Tech Vanguard's human resources department had never bothered to update the title.
After six months, his supervisor began assigning him ordinary janitorial duties because the maintenance department was short-staffed.
Jamal had accepted the change.
He needed the job.
His mother needed treatment.
The distinction between cleaning a laboratory and maintaining its equipment mattered very little when the alternative was missing a payment for her medication.
Still, he had not stopped being an engineer in the way he thought.
He noticed how machines sounded.
He watched how people approached problems.
He read technical manuals during his lunch breaks and spent his evenings working through online lectures on thermodynamics, signal processing, and control systems.
When an engineer discarded a damaged component, Jamal sometimes asked permission to examine it before disposal.
Most refused.
A few agreed.
He had accumulated a modest collection of parts in a storage cabinet in his apartment, where he practiced identifying failure patterns and understanding why certain components wore out faster than others.
He did not pretend this made him equivalent to a research team.
He knew the limits of his experience.
But he also knew that careful observation could reveal questions worth asking.
And for six weeks, one question had bothered him more than anything else.
Why did the engine shut down after exactly fourteen minutes and thirty-seven seconds?
The interval was too consistent to ignore.
A random mechanical failure might occur at different times under different conditions. A software problem might repeat, but its behavior would usually depend on the state of the system and the sequence of events that preceded it.