THINGS WORTH WANTING / PRESENCE & ROOM DESIGN
Aqara FP2: You Shouldn’t Have to Wave at the Lights
A room should not need a performance to know you are still there.
By Dr. Robert Urban · Researched editorial feature · September 2026
You sit down to read. A few minutes later, the room decides you have left.
The traditional response is a small, undignified dance: one arm lifted from the chair, perhaps a foot added for emphasis. You are no longer simply enjoying the room. You are submitting evidence of continued existence.
Aqara’s Presence Sensor FP2 addresses the distinction behind that familiar irritation. Entering a room and remaining in it are different events. A useful automated home needs to understand both.
The FP2 deserves consideration for that reason. Its appeal is the possibility of making a room less demanding, provided the owner is willing to do the configuration that such apparent simplicity requires.
Remaining is different from arriving
According to Aqara’s FP2 product information, the sensor uses millimeter-wave radar for presence detection, supports configured zones within a room and includes a light sensor. It requires continuous power. It does not need a camera to perform its presence-sensing role.
Those capabilities create a useful design opportunity. A reading chair and a doorway can belong to the same room while calling for different behavior. Passing the chair is not the same as sitting in it. Getting up briefly is not necessarily a request to extinguish everything.
The distinction should guide the purchase. If a basic arrival-triggered light already works reliably in a short passage, a more elaborate presence sensor may add little. A room used for quiet work, conversation or reading presents a more interesting problem.
This feature examines the design and ownership implications. It does not claim a measured detection rate or firsthand testing in a specific floor plan.
Design the behavior before drawing the zones
Begin by writing one sentence about the desired result. “Keep the reading light available while this chair is occupied” is a useful brief. “Make the entire house smart” is a shopping expedition with no stopping point.
Then identify the circumstances in which that result would be unwelcome. Someone may sit in the chair during daylight. Someone may prefer the lamp off while using a screen. A second person may be sleeping nearby. A sensor reports a condition; it cannot settle every question of household courtesy.
That is why an automation should combine the necessary context and preserve an understandable override. Presence may be one condition, available light another, and the household’s chosen mode a third. These are proposed design principles, not a promise that every ecosystem exposes identical controls.
Keep the first version small. One room, one useful action and a clear way to suspend it will teach more than a dozen linked rules whose effects are hard to separate. Complexity is easiest to introduce before anyone has lived with it.
A zone should describe an activity
A floor plan encourages tidy rectangles. People encourage less tidy behavior.
The useful zone is the area associated with an activity, with enough thought given to approaches, exits and adjacent use. A dining area might include the chairs when pulled back. A desk area should reflect the person actually working there, not just the outline of the desk.
Observe how the household moves before treating a map as finished. Does someone lean into the area to collect a book? Does a route between two rooms cross the intended boundary? Could one person leaving a shared area trigger a change that bothers someone who remains?
These questions are especially valuable because they turn configuration into a reviewable decision. Instead of saying “the sensor is acting strange,” an owner can ask whether the map represents the behavior they intended to recognize.
Placement, room geometry and the current setup instructions deserve attention. Aqara provides an FP2 support page. Use the guidance for the exact model and operating mode, then evaluate the actual room. Manufacturer capability statements are not evidence that a particular installation needs no adjustment.
A reading chair is not a corridor
Aqara advertises up to 30 zones, but the useful design exercise can start with two: the route through a room and the area around a reading chair. Its product guidance specifies wall mounting for zone positioning. Decide where continuous power and that mounting arrangement will work before planning behavior around an idealized floor plan.
In this illustrative brief, a person crossing the room should not activate the reading lamp. Occupancy in the chair area can permit the lamp to come on when the room is dim and the household has enabled that behavior. These are proposed rules to implement in a compatible controller, not a claim that the FP2 ships with this scene.
Departure needs a different rule from arrival. Allowing a grace period before switching off can accommodate someone collecting a book and returning. The interval should follow the household’s tolerance: an extra minute of light may be preferable to repeatedly interrupting a reader. There is no universal timer that can make that judgment for every room.
Manual off needs its own answer. If a reader turns the lamp off to watch a screen, a rule that immediately turns it back on has misunderstood the person. Where the controller supports it, an explicit automation-enabled condition can preserve that choice until the household deliberately resumes automatic operation.
Finally, test a second person crossing the room while the first remains seated. Record the unwanted result precisely: an arrival triggered the wrong lamp, a seated reader became absent, or a manual choice was overwritten. Each observation points to a different part of the configuration. Redrawing every zone at once makes the cause harder to find.
Test the moment when nothing happens
Most demonstrations favor arrival. Walk in, light comes on, everybody understands the point.
For a presence sensor, the more revealing test is quieter. Sit where you normally sit. Read. Work. Have another person enter and leave. Leave briefly and return. Try the situation with daylight and without it. Check what happens when you deliberately turn the light off yourself.
You are looking for predictable behavior, not a dramatic demonstration. An automation that is impressive nine times and irritating the tenth will train people to distrust it. The remedy may be a different placement, a simpler rule or a longer delay before an absence-based action. Change one thing at a time so the result remains interpretable.
Make the cost of a mistaken action part of the design. Turning on a decorative lamp unnecessarily is usually a different inconvenience from turning off the only useful light while someone is working. Those actions need not use equally aggressive conditions.
The house should remain easy to use while the automation is being refined. Preserve familiar manual controls. A guest should be able to ask for light without learning why the owner drew a zone around the armchair.
Local operation is a whole-path question
Aqara describes local automation support with specified compatible ecosystems and says an Aqara hub is not required for the FP2 itself. That does not establish that every combination of sensor, controller and target device will work without internet access.
Ask about the complete path. What receives the presence information? Where does the rule run? How does the command reach the lamp or switch? A cloud-dependent step can remain in a system whose sensor is capable of local communication.
Home Assistant documents a HomeKit Device integration for connecting compatible accessories. Aqara identifies that route for FP2 use with Home Assistant. Pairing requirements matter, including the controller to which an accessory is already paired. Follow the current documentation rather than assuming that two familiar logos guarantee an effortless handoff.
If this becomes a larger project, Home Assistant Green raises the next ownership question: who maintains the logic after the initial enthusiasm passes? The presence sensor provides information. Someone still has to decide what the home should do with it.
No camera does not mean no conversation
Camera-free sensing can be attractive in private living spaces. It still deserves a household conversation. Explain what has been installed, what it controls and which information is available to authorized users.
Avoid making a comfort automation sound like a comprehensive account of a person’s activity or wellbeing. Presence information has a specific purpose and limits. For this feature, the compelling use is ordinary room behavior, not health monitoring or an assurance that someone is safe.
That is a useful companion to the Palladiom discussion of daylight and control. One product changes the light entering a room; the other can inform when an action is appropriate. Neither removes the need to think about the people using the space.