Since Tado limited their API rates — and they were practically the only “smart” thermostat on the market — I started thinking: what if I build my own? And more importantly, what if I can do it better?
First of all, why ask here?
Because you are the kind of users who do more than just set the thermostat once and forget about it.
What is the plan?
I want to create a universal gateway. This gateway would be the only device physically connected to the heat pump, boiler, or central heating system.
Thermostats themselves would only connect to the gateway for power. The actual control logic could come from anywhere: Home Assistant, MQTT, Bluetooth, an API, or any other system.
In other words, the thermostats could be anything — physical devices, software integrations, or custom automations.
My goal is to do this properly and, ideally, make it open source. I would much rather build something that lasts 10+ years than lock it behind subscriptions or paywalls. Probably the worst business plan ever, I know.
Until now I’ve mostly kept the idea to myself, mainly because I didn’t want some company to take the concept and turn it into a closed commercial product.
So I’d like to ask:
Would you be interested in something like this?
If so, what features would you want? How would you prefer to control your heating?
Proposed Features (early concept)
Universal heating gateway
One device connected to the boiler / heat pump / central heating system.
Works with any thermostat source
Control can come from Home Assistant, MQTT, API, Bluetooth, or custom integrations.
Local-first design
No cloud required. Everything runs locally.
Open source
Hardware and software intended to be open for transparency and long-term support.
Multiple heating protocols
Planned support for OpenTherm, 0–10V control, and simple relay control.
Safe and robust control
The gateway is always the single device controlling the heating system to avoid conflicts.
Flexible thermostat options
Physical thermostats, dashboards, automations, or sensors can all act as the “thermostat”.
Future-proof design
Designed to work for 10+ years without subscriptions or vendor lock-in.
I’d love to hear as many ideas and opinions as possible so I can make the system as universal as possible.
Great initiative! To clarify what you’re proposing: are you building an open source Tado server/proxy that lets existing Tado devices work locally (bypassing their cloud), or a full DIY hardware gateway + thermostat replacement?
My priority (and I suspect many others’): A local Tado server/jailbreak that unlocks our existing hardware. I dropped €1000+ on Tado TRVs less than 2 years ago, drained my central heating system multiple times to replace fittings for the installation, I’m not ripping them out for new hardware.
But their API limits and cloud dependency? Unacceptable enshittification. A local server that emulates their cloud APIs would be perfect keep the TRVs we own, ditch their app/cloud entirely.
I want to find a way to (professionally and safely) control the heating unit. For what it’s worth, I want to support thread, matter and zigbee. Also wifi an bluetooth. Which should mean there is a way to connect tado stuff locally. (and for what its worth most other thermostat, step by step)
Still in the designing / drawing schematics progress. Looking into chips and I don’t think ESP will cut it, so it needs more precise programming and system design. I will keep everyone posted!
I want this to be as universal as possible. So support for all kind of TRV, Thermostats. Including TadoX. But I am still in the drawing board and looking for suggestions on what not to forget!
All suggestions are appreciated, from the amount of LED to connectivity to chip.
@douwehoeben - that would be amazing. Will follow this closely, me as well I dropped a shitton of dough into tado x, just to be royaly fucked. Anyway…
@matrover - Tado X working fully locally via the HA thread network? Please do tell…
Most of things are not exposed.
Im currently using tado x - still days when I am being signed out, due to api being hit, not to mention the updates are 2700 seconds apart (45 min).
I’m also very interested in this project. Particularly the ability to remotely control commercial grade zone actuators such as Siemens SSA161.05HF where it takes a 0…10V control signal. When it comes to infrastructure I’m willing to spend big bucks for quality and reliable items, but hard wiring is also a must for me. In this case, low voltage or Ethernet cables seems to be the best options.
The zone actuator I’ve mentioned goes for about 100 to 120 euros a unit. Yes, it is expensive, but they are also rated for a much longer life, commonly between 10 to 15 years, so it’s worth it in the end.
I would suggest taking a look at how HeatMeister is doing this using ESP32-C2 MCUs, it can do a lot more than what you think. This product also has a huge forum where people are sharing their own ways of making radiator fan mounts.
Definitely sounds like an interesting concept! I’ve been wanting to make a similar thing as I’ve got Sonoff smart TRVs, but I want them to be controlled by separate zigbee thermometers in each room, not their local thermometer (like yes, behind my sofa where the radiator is will obviously get up to temperature faster, but I want to use actual room temperature).
The ability to enable the heating via a smart relay (I’ve got a Matter enabled relay connected to my boiler) if one room is low…
Would be amazing if something could actually tie those three separate things together on a non-cloud-dependent system!
Hi Yerwol:
Forgive my ignorance, but there’s something in your comment I don’t understand.
If you’re using Zigbee thermometers, why would you put one behind the sofa? My Zigbee thermometers are battery-operated, and I can put them wherever I want.
