Why Presence Lighting Turns Off While Someone Is Still in the Room

A sensor can miss a person who is still in the room

A Reddit kitchen-lighting discussion describes two passive infrared sensors turning lights off while the owner sits still. The same owner reports that a trial mmWave device responded to blinds and dogs while missing the intended location. This single account illustrates both missed occupancy and unwanted detection; it does not prove one technology always succeeds.

First compare the sensor’s occupancy history with the time the light turned off. If it reported vacant before shutoff, investigate sensing and timing. If it still reported occupied, inspect the command history for another routine. Competing commands are a separate problem covered in the related guide.

Source: Kitchen lights time out while seated; owner compares PIR and mmWave

Choose detection for the activity, not the label

Passive infrared, or PIR, sensors respond to changes in infrared energy across their sensing zones. A person walking through coverage is a different target from someone reading with little movement. Millimeter-wave radar, commonly called mmWave, can detect smaller movements, including breathing-related motion under suitable conditions. Neither technology literally proves that a room contains no person.

Aqara’s manufacturer guidance describes this distinction. Use the actual product’s mounting, range and sensitivity instructions; furniture, viewing angle and the occupied position affect the result. Test the chair, desk and work counter rather than demonstrating only someone walking past the doorway.

Source: Aqara: Presence and Motion Sensor Detection

Address false occupancy as well as false absence

Aqara’s FP2 manual identifies fans, moving curtains, plants and similar moving objects as possible interference sources. It provides edge and interference-zone settings for that model. A sensor that detects a moving fan as a target may keep a room lit after people leave; raising sensitivity is therefore not a universal repair.

Map the actual occupied area and moving objects, then change one supported setting or placement at a time. Keep a record of both occupied and empty-room tests. Product-specific exclusion zones should not hide an area where someone legitimately sits.

Source: Aqara FP2 Manual: Edges and Interference Sources

Use a vacancy policy suited to quiet activity

Treat an absent reading as input to the rule, not an instruction that must immediately black out the room. A suitable delay, additional evidence or an obvious manual hold can make quiet use practical. The correct delay is an agreed household choice, not a universal number.

For example, test a reading chair through a normal quiet session, then test how the lights behave after everyone leaves. Explain when a manual hold expires and how normal operation resumes. Do not compensate for a blocked sensor with an indefinitely extended timeout.

Test the occupied positions and an empty room

Record sensor state, automation decision and final light state as three separate observations. Repeat representative tests after furniture or sensor placement changes. A manual app command proves that the light accepts commands; it does not verify occupancy detection.

Retain an accessible ordinary control and a short record of the accepted settings. If sensing remains unsuitable for a particular activity, choose a simpler supported control policy for that room rather than treating more automation as the required outcome.

Put it into practice

A decision table to keep handy