What's the weirdest cause of a machine failure you've come across?
Heard about a production line that kept shutting down at roughly the same time every afternoon. Maintenance checked just about everything before someone realized the sun was hitting a photoelectric sensor through a window.
A simple shade fixed what had been driving everyone crazy for weeks.
Classic receiver saturation. Most industrial photoelectric sensors use modulated (pulsed) emitters so the receiver can reject steady ambient light, but direct sunlight or a reflection off a shiny surface can still drive the receiver into saturation, and the output then reads as blocked or unblocked depending on the mode. The datasheet ambient light immunity rating (in lux) tells you how much margin you actually have.
Things that usually fix it for good: mount the receiver or reflector so it does not look toward windows or skylights, add a hood or shroud, use polarized retroreflective sensing with a corner-cube reflector so stray glare is filtered out, or switch to background-suppression diffuse sensing where the geometry allows it.
The broader lesson: when a fault repeats at roughly the same time of day, log trip timestamps and compare them against sun angle, shift changes, HVAC cycles, washdown schedules, and large motor starts nearby. Time-correlated faults usually point to an environmental cause rather than the device itself.
Things that usually fix it for good: mount the receiver or reflector so it does not look toward windows or skylights, add a hood or shroud, use polarized retroreflective sensing with a corner-cube reflector so stray glare is filtered out, or switch to background-suppression diffuse sensing where the geometry allows it.
The broader lesson: when a fault repeats at roughly the same time of day, log trip timestamps and compare them against sun angle, shift changes, HVAC cycles, washdown schedules, and large motor starts nearby. Time-correlated faults usually point to an environmental cause rather than the device itself.