A friend at work mentioned they are participating in an investigation where a series of battery powered sensors have been installed in his house that then communicate via a Raspberrypi with a cloud back-end. The idea is that the participants are monitored and given recommendations by the system to improve their health.
The thing I found most interesting is that they have installed a series of battery powered sensors around his house that detect movement and also the height of the object causing the movement. The system then can identify who is triggering the movement and act accordingly.
I thought this sounded like a really nice thing to be able to integrate into HA.
Does anyone have any idea about the hardware they might be using and if it would integrate well into HA?
They are not using image recognition as the sensors don’t contain cameras. My friend said he asked the guy who did the installation if they had thought of using cameras, apparently they didn’t want to use cameras as otherwise the sensors would use too much power to be battery powered.
I will ask my friend if he can try and get more info about them.
My 2 cents… ultrasonic might be able to do the trick. The main questions to ask would be what the accuracy is? how much time does the subject need to be under the sensor? where is the sensor placed? and the area covered by the sensor because the location of the person might affect.
Based on experience these might be a bit iffy and not too reliable unless they have more accurate sensors or an array of sensors. Which is why I’ve seen these applications move towards image recognition (specially using stereoscopic cameras).
Apparently the detection is pretty fast as it detects people without them being still. The sensors are at about 1.8m height. One in the bedroom one in the lounge on in the bathroom and one in the hallway.
I can see this being advantageous for people who don’t want cameras (especially in their bathrooms) but who want to track who is where in the house.