Which sensor for baby presence detection in a room?


I bought FP2 Aqara detectors for 3 rooms in my apartment. Overall, I am very satisfied with the operation and I have done well with the automations I wanted with Home Assistant.

On the other hand, it does not work as I would have liked for my baby’s room: the sensor only detects his presence very randomly. I would like to program automations depending on whether the child is in the room or not: leave the nightlight on, play a lullaby…

If I can’t rely on the FP2 sensor, which mainly detects the presence of adults, is there another way to detect the presence of my daughter when she sleeps in her crib? Do you have any recommendations for sensors that would work with Home Assistant?

As it is, in her room, I have installed the following HomeAssistant compatible devices: FP2 sensor, an Aqara temperature sensor and a Xiaomi Mi Speaker. I have also a baby camera “Lollipop” which is not compatible with Home Assistant.

Not being very handy, I am ideally looking for a solution that does not have to be self manufactured.

Many thanks for your help

Normal motion sensors don’t detect small movements that well.
Small movements require mmWave or similar, like Everything Presence One Kit – Everything Smart Technology

I have no idea how a mmWave sensor will affect a baby, but EU’s car safety programs wants to use them to detect children in the car.

How about our sensor? Affix it to baby’s mattress and sensor will update “status” with “in-bed”, “empty” or “movement detected” every second. Very accurate :slight_smile:

1 Like

Sounds interesting, but it all depends on price and time to market :slight_smile:

Normally, Aqara’s FP2 detector is an mmWave detector, but this is not enough to detect the baby :frowning:

Maybe I’m going to say the obvious now, but usually the child moves quite actively and your sensor will definitely detect it. And when the child goes to bed, it is enough to put a pressure sensor under the mattress and this will be enough for you to know for sure that the child is there.

I absolutely agree with you. When the child will be older and she awake, she will be active and will be detected by the FP2 in my opinion.

Today, she is basically in her room sleeping, and moves very little, so the FP2 does not detect her. A pressure sensor under the mattress would be great, but I haven’t found any for sale. Only DIY tutorials which I find a bit difficult as a non-do-it-yourselfer with little time to spare.

Have you tried aiming the sensor straight into the crib?
I think any mmWave sensor will be able to detect the presence of the baby in the crib,
but you will probably need a second sensor for the presence of the room.

1 Like

Indeed, I tried to install the FP2 Aqara sensor in the cradle, but the detection is not reliable at all. It does not detect the baby most of the time.

I was thinking of another alernative: when I put two temperature sensors under the mattress with one under the baby, and one further away, I notice about two degrees of difference. Is it possible to make automations that will compare the temperatures of two sensors and if the difference is greater than or equal to two degrees, consider that the baby is in his bed?

I am interested in any other idea :slight_smile:

weird i have mmwave sensors and they detect breath movement.

You can create a sensor template

Thank you for your answer.

Maybe I should try another mmWave sensor in this case, just for the cradle? But I must admit that I would like to be sure of the result before investing again.
Or is it the baby camera that interferes?

However, the FP2 indicates to be a mmWave sensor : https://www.aqara.com/eu/product/presence-sensor-fp2

I don’t think that camera influences the detection.
Usually the problem with mmwave sensors is false positives,
Swinging curtain, fan, washing machine, etc.

Does the FP2 have sensitivity adjustment?

I use the zigbee model, for wall:

This sparked my curiosity and I searched around for academic publications on health effects of low power mm waves on infants. Results are inconclusive. Low powered mm wave sensors are a new technology and there’s not enough data on long term exposure yet. Effects are mostly heating related, but some research seem to indicate that penetration into lower skin layers and into the eyes is possible and can alter heat-dependent biochemical processes.

It’s most likely harmless on adults. But exposing a baby to targeted and constant long term RF (unlike that car thing, which is only short term), in its sleep (where lots of vital biochemical processes happen), while its organism is in full development and may be very sensitive to chemical alterations induced by RF exposure… I’m not sure this is such a good idea to be honest. Especially if its only about convenience for the parents and there’s no medical indications. Just my own opinion ofc.