He enlarged the diagram and compared it with an archived set stored in a local reference folder. Several components carried unfamiliar numbers, but the underlying control relationship was one he had removed from an early design because it behaved badly under certain conditions. A revision that had once required a long explanation had been reversed without the original warning attached. Ethan searched for the validation history, then for the approval record that should have accompanied the change.
There was no complete record. What he found instead was a summary stating that the newer implementation had passed simulation, followed by a proposed test range ending beyond the narrow band that concerned him. One figure held his attention long enough for Noah to notice. Forty-one point six hertz.
"Something wrong?" Noah asked. Ethan closed the file and looked through the glass wall at the large prototype sitting under white work lights. "Who approved the latest architecture?" he asked. Noah said the decision had come through executive engineering, but he could not identify the person who had signed the detailed validation.
Ethan knew what it felt like to discover an old mistake disguised as a new idea. It was not necessarily fraud, and it did not mean the designers were careless people; engineering teams often inherited decisions whose reasons had been forgotten. But the absence of the original restriction meant somebody had decided the design was safe without knowing what had once made it unsafe. He asked Noah to preserve the version history and told him not to let anyone run the unreviewed configuration at the suspect frequency.
"Shouldn't we tell the lead engineers?" Noah asked. "We will," Ethan answered, already reaching for a notebook. He wrote the first observation in the plain, patient language he used when preparing a report he might later have to defend. The problem was serious, but at that point he still believed the company would be grateful to know about it.
Thirteen days after Ethan first recognized the altered architecture, Northstar prepared a demonstration for three representatives from Washington. The test was scheduled for a Tuesday morning, and the company had decorated the corridor outside Laboratory Three with banners carrying words such as reliability, innovation, and trust. Security officers guided visitors through glass doors that had been polished twice before sunrise. Inside the laboratory, a group of technicians argued quietly about whether the test schedule allowed time for a final review of the unfamiliar control pathway.
Clare Bennett arrived at 10:17 with Vanessa Reed and six senior engineers. At thirty-six, Clare was the youngest chief executive in Northstar's history and the daughter of the founder whose name remained on the headquarters building. She had spent years watching older executives congratulate her on being prepared, then call her father afterward to confirm what she had said. She had learned to speak precisely and prepare twice as thoroughly as anyone who might challenge her, a habit that made her formidable and sometimes left her unable to tolerate being surprised.