Intermittent USB and Display Drops in an Open Office
Over three weeks, four users in one open-plan office report that their USB keyboards and mice freeze for a second or two, and that the second monitor on their docking stations occasionally goes black for about two seconds before returning. All four use the same laptop model and the same 90 W dock. Two sit at desks 4 and 5, which share a power strip under a shared desk; the other two are across the room on different circuits. New LED lighting was installed on the floor above two weeks before the reports started, and a microwave was added to the adjacent break room. Users say it is worse in the afternoon. Swapping one affected user onto a known-good spare dock did not change the symptoms. No other floors report anything similar, and the devices are otherwise fully functional.
1. You have limited spare parts and one afternoon. Design the sequence of tests and state what each test rules in or out.
Read the worked answer
Order the tests by cost and by how much the result narrows the hypothesis space.
- Confirm the symptom is still present under observation. Watch a user reproduce it. Many "intermittent" faults are actually a specific trigger (plugging in a second device, closing the lid) and observation costs nothing.
- Swap the dock - already done, and it ruled out a single defective dock. Because the fault survived a known-good dock for that user, dock failure is now a weak hypothesis. Do not spend more spares here.
- Move one affected laptop to a known-good desk away from desks 4-5 and leave it for an hour. If the fault follows the user, it is the laptop or its drivers. If it stays at the desks, it is environmental or shared-infrastructure. This is the single highest-value test and uses no parts.
- Swap the USB-C or Thunderbolt cable and the monitor cable on one desk. Cables cause exactly this class of brief, repeatable dropout and are the cheapest thing to replace.
- Check driver and firmware versions against the dock vendor's release notes. A dock firmware update or a graphics/Thunderbolt driver mismatch can produce identical symptoms across an entire identical fleet, which fits the fact that all four users share a model.
- Time-correlate with the microwave and the afternoon pattern by logging occurrence times for two days. Do not act on the correlation until you have the log - the 2.4 GHz band is a plausible culprit for the USB wireless peripherals, but only if the peripherals are wireless, which you must verify first.
Each step answers a different question; running them in parallel would tell you that something changed but not what.
2. The fault follows the laptop when it is moved, across all four users. What does that change about your diagnosis, and what do you do next?
Read the worked answer
It eliminates the largest branch of the tree. Shared infrastructure - desks 4 and 5, the power strip, the room's electrical environment, the microwave, the LED lighting upstairs - is now ruled out as a cause, because the fault travelled with the machine. It also devalues the lighting and microwave theories, which were correlations built on timing rather than on evidence.
What remains is a fleet-wide common factor: same laptop model, same dock model, same driver and firmware baseline. Four identical machines failing identically is the signature of a software or firmware defect, not four coincidental hardware failures.
Next steps:
- Capture the actual driver and firmware versions from all four machines and diff them against a working machine of the same model - the odd one out matters more than the majority.
- Check the laptop and dock vendor's known-issues list for this model and firmware combination, and check whether a recent update shipped to these four before the reports began.
- Look at event logs and, if available, dock firmware logs for the exact timestamps of the dropouts, then compare with update installation dates.
- If a firmware or driver release fixes it, pilot on one machine for a full day before rolling out.
Also revise the ticket: the original theory was environmental, and the record should show why that changed, so the next technician does not repeat the investigation.
3. The microwave theory is popular with the users. How do you handle it honestly without dismissing them?
Read the worked answer
Treat it as a hypothesis with a testable prediction rather than an opinion to be accepted or rejected.
Explain the mechanism in one sentence: 2.4 GHz cordless and Bluetooth-class devices share spectrum with microwaves, so a leaky unit can cause brief interference during operation. That is real, and it is why the theory is worth one cheap test rather than a flat no.
But state the prediction clearly: if the microwave is the cause, dropouts should cluster tightly in the minutes it runs, should not follow a laptop to another floor, and should affect wireless peripherals more than wired ones. Your evidence already contradicts the second prediction. So run the first test anyway - ask users to note the time whenever it happens for two days and compare against break-room usage. A null result retires the theory with data instead of authority.
Communicate the outcome to the group, including what you ruled out and why. Users who supplied a plausible theory and were told "no" become silent; users whose theory was tested and disproved become collaborators, and they will keep reporting the timestamps you need for the next fault.