That would be odd, generally non infrared spectrums will not affect it, UV is on the far opposite end of the light spectrum from what the sensor uses, I would be more inclined to suspect maybe heat from the unit or magnetic interference, infrared is much more like heat in fact with the old osmolator, being near a heater could mess up the sensor but only heaters with a visible nichrome coil, those old bathroom heat lamps and chicken coop heat lamps or ceramic heaters are basically infrared light bulbs. I can't rule out that the window has a filter that could have some conversion to infrared (unlikely) or a red/orange pilot light to let you know it is on could be a cause. In the past with the old unit we sometimes saw issues with certain ballasts that would transmit interference and the wires would carry it like an antenna, but in this case, a radio would also pick this up and that kind of went away as an issue with compact flourescents no longer being a factor in the hobby.
