I have my sensor setup in my master bathroom and it’s working great but sometimes the light bounces on and off and gets stuck in when the room is empty. It’s parallel to a giant mirror, has anyone had any issues with reflection ghosting and found a solution?
1 Like
I think this is possible, and you can use AI spatial background learning to learn about environmental noise
1 Like
Hi @RipVanWilliams, which sensor are you using exactly? I assume it’s one of the mmWave presence sensors like the Aqara FP1 or FP2.
For context: mmWave stands for millimeter wave radar. These sensors constantly emit very high-frequency radio waves (usually 24 GHz or 60 GHz) and analyze the reflections to detect even tiny movements – like breathing – so they can tell if someone is in the room even when they’re sitting still. That’s the big advantage over PIR motion sensors, which only trigger when there’s obvious movement.
The catch is: a mirror with a metallic coating behaves very differently from a normal wall.
- Normal wall (stone, drywall): Partly reflects, partly absorbs, and scatters the signal.
- Mirror with metal coating: Reflects the radar waves almost completely and very cleanly, like a radio “echo” – just like it reflects light.
If your sensor is parallel to a large mirror, it might be “seeing” ghost signals from its own reflections, making it think there’s still movement when the room is empty.
Possible fixes:
- Angle the sensor slightly so it doesn’t point directly at the mirror.
- Move it away from direct line-of-sight with the mirror.
- Fine-tune the detection zone in the app so the reflected area is excluded.
mmWave is great – but in a bathroom full of shiny surfaces, sometimes it’s too good at seeing things that aren’t really there. Hope it helps…
Always remember: “If your car is smarter than your home… you’re driving into the future while living in the past” 
4 Likes
I can confirm this 100%. I had an FP2 sensor in the bathroom above the door, and opposite the door was a mirror. I had a lot of problems at first until I came up with this theory myself. I then placed the sensor above the mirror, re-trained it, and had little to no ghosting. Furthermore, I discovered that the radar waves can also pass through thin walls and/or doors. Therefore, I recommend marking the exit on the FP2’s map. This seems to help the FP2 better understand the room. I therefore recommend not setting up the FP2 opposite mirrors, windows, or glass doors. This is the experience I’ve gained with five of these sensors.
3 Likes