Elias was a structural engineer assigned to a routine retrofit of the grandstand roof at a mid-sized arena that primarily housed a local hockey farm team. Although Elias was experienced, this was one of the larger projects that he had worked on, and he was both excited by the new challenge and somewhat anxious about gaps in his knowledge. The project was relatively straightforward, but availability of some materials and few other issues caused the team to already be three weeks behind schedule, and the pressure from the city council to finish before the season opener was immense.

One place where no expense had been spared was the monitoring system for the roof’s tension cables. It was state-of-the-art. It provided real-time data on load distribution, thermal expansion, and vibration. Every morning, Elias checked the dashboard, and every morning all indicators were green. The load factors were always well within the safety margins. On Tuesday, Elias was on a routine site walk-through. The day was particularly windy, but everything seemed to be handling it well. However, at one point, Elias felt a vibration in the handrail of the catwalk that didn't seem quite right. He paused, and noticed that there was also a low-frequency hum that was syncing with the wind gusts.

"It’s just resonance," the Site Foreman told him, checking his watch. "The sensors are flat. If there was an oscillation issue, the accelerometers would be screaming. We have to pour the concrete for the footing today. We can't stop for a hum."

Elias knew the foreman was right, but when the walk-through was completed, he dug into the monitoring system data again, but everything was clean. The million-dollar system that had just been calibrated was more trustworthy than his hand on a railing. With a hundred other tasks on his plate, Elias was tempted to move on. But still, he couldn't shake the feeling that something was off, and he wanted to learn more about this system anyway.

Risking further delay, and putting the concrete pour permit on hold, Elias went back to the trailer and pulled the schematics for the sensor installation. It took a while, but then he had it. The accelerometers were placed at the nodes of the truss structure—the points where vibration is naturally lowest—rather than the antinodes, where movement is greatest. The sensors weren't broken; they were just blind to this specific frequency of vibration.

Now, he really wondered what was going on. Elias climbed back up with a handheld vibration meter. He measured the mid-span movement. It was dangerous—bordering on critical. The "hum" was the sound of the steel fatigue setting in.