Next generation will support USB C (now is micro USB) and maybe uSD card.
There were also plans to support the sniffer of 433MHz / 868MHz / 2.4GHz for non-standard communications (like gate remote controler), but due to market situation will be a separate device.
But my question is on a different topic, I design devices on a daily basis and I thought that there is no dedicated devices for HA.
I think Home Assistant has a lot of potential that can be used, so maybe try to create a series of devices that HA supports Plug&Play? Dedicated for HA.
We would create ready-made integrations so that the user would not have to write any scripts.
We are currently testing pairing/assigning a device using NFC, the central station already has a reader and each device would have a TAG. This would only allow the device to be brought closer to the central unit to assign a new device to the system (it does not even need to be powered). This can also be done from the phone application also.
I’ve already tested voice recognition (local without network connection) in different languages and it works pretty well (dedicated device for this).
What do you think about it?
If you like the idea, what devices would be the most useful at the beginning?
These can be existing solutions adapted to HA or new concepts.
But this doesn’t run home assistant itself (like the home assistant yellow) - right?
By the looks of it some technologies (like Matter, Thread, ZigBee) are already part of the yellow.
Other stuff like BLE, CAN, RS485 (and many more like RF, IR, NFC, RFID, etc.) can be easily integrated with (very) cheap esp devices and seamlessly integrated with esphome.
What “niche” would your device handle @adi-mistrzu? Something like “satellites” which are scattered around the house and capturing the signals which are then forwarded (via wifi or lan) to the home assistant server? Like the recently introduced bluetooth proxies?
Thanks for answer.
Home Assistant already work on it (also supervised).
Gateway will be smaller and should be cheaper (it is not based on ready modules and due to its size it is cheaper to produce).
With esphome, in my opinnion, the basic problem is the need for reprogramming.
Many amateurs are unfamiliar with it and are unable to do it.
Releasing devices specifically with HA support from the beginning could increase the popularity of the Home Assistant among people who are not familiar with programming or electronics.
The idea was for devices (even similar to those currently on the market, such as switches, but not only) that were easier to connect with HA (like via NFC). Without requirement any reprogramming, scripts etc.
So it is kind of a yellow replacement or “competitor” then? What SoC is the system based on? Does it perform “sufficiently”? Also I can’t spot any antennas? No usb - so no way to extend storage (or other things) and once the eMMC/NAND module is dead the nice little device can’t do no more? (just recently received two “stick” pc’s with dead internal flash modules - luckily they have usb so it was possible to still make use of the hardware with a external ssd in that case).
That really depends - if you just want a esp32 to act as a bluetooth proxy you don’t even need to have esphome installed to do that - it’s easy enough to open that site and install it directly onto your esp:
In general Home Assistant is (still not) a plug’n’play solution but quite the opposite in most senses. That is not essentially bad as it gives lot’s of freedom to it’s users.
I’m not quite sure if this will be a thing. The leraning curve is quite steep (“with freedom comes great responsibility”) and actually there is already a lot’s of limited (and therefor much more simple) solutions out there users could choose from if the golden cage is enough for one
The problem I see here is that the gateway needs a way (software/integration wise) to talk to ha and on the other hand all the devices connected to the gateway need to be already working with ha. So if a user has this gateway but it’s xyz-Radio device does not work the frustration will be big when before the claim was “works out of the box”.
I think the ground work for such hardware needs to be done in software before. It needs “proxies” (like the bluetooth proxy) for all the protocols so than hardware like yours can really work “plug’n’play” and forwarding all information transparent to home assistant (which then is in charge for detecting/decoding/knowing etc. the “end” devices).
I use NXP i.MX8M Nano 4 core, I have not noticed any problems with the lack of computing power (it is the main router with DHCP server for me also).
Antenat is buil-in (due to regulations, it must be inseparable anyway), there is externat uSD connector, but will be moved to more accessible place.
USB support OTG, so if necessery do connect external device, then I connect the adapter. The next version will be USB-C so it will be even easier.
Only SSD M.2 is not supported.
I fully agree, I’m currently working on making the software work as P&P.
Releasing this type of device with “raw” HA would not work, it would be just another platform.
Therefore, by launching such a device, the HA would already have planned support for easy device adding (without losing current capabilities).
So profit from hardware would be invested in HA software development and P&P devices.
Therefore, I would like the community to ask what it looks like from their perspective, or maybe my misconception is somewhere.
