"Smart" climate function

I move from one apartment to another. Before I’ve the Nest thermostat linked to the boiler. But now, it’s a common boiler, and each apartment has several radiators with is own thermostat. So I replaced the thermostat by an EQ3 Bluetooth.

But how to make it “smart” ?
I really like the function on the Nest, It can anticipate how much time it’ll take to reach the chosen temperature (depending on the outside temperature,…)

Is there a way to install those “smart function” ? With Home Assistant you could combine as many sensors (like I have 5 temperature sensor in MQTT, and you can place them in your own way), without the Big Brother worry and without depending on a specific brand or better where you can integrate your own DIY things (like with MQTT).

Anyone has ideas to predict the temperature reaching time, estimate Thermal inertia of a room … ?

I have external temperature sensors, room temperature sensors and electronic valve for the heating for each room.
I would like to know how much time it needs to reach the wanted temperature.

So build a model and collect data
How cold was it outside, how warm was it inside, were all the doors open so that heat could migrate evenly, were all the windows closed, how hard was the wind blowing, was there snow piled up against the walls or windows, how long did it take to get 1 degree higher
And then the next degree (it won’t be linear as radiation increases with differential)
When you have your data for a range of external temperatures and internal start temperatures and conditions you will be able to build your model for ‘you’ living in that apartment. Somebody else may be completely different as they always close this door and that door. Or they have different basel electrical loads as they have different appliances.
It’s not easy, as nest was ‘supposed’ to be ‘self-learning’, most people I knew who had one, turned those bits off after 8 months or so. Also ‘some’ (nest systems) adjusted the boiler water temperature according to outside temp, you will need a special api to do that and the right bit of kit. Though setting the temperature too high defeated the condensing aspects and massively reduced efficiency.
You pays your money you takes your choice.

Maybe HomeAssistant is not just to turn on and off some light ?
I’m sure it interest other peopel and I made this post in the “feature request”.

If I had more knowledge in deeplearning, I would try.