Problem creating device profile

Good morning

I have recently started using Smart Things with the aim of directly connecting an ESP8266 to the Smart Things App.

As I ran into a problem, I closed my account and started again (I had unfortunately done this before reading an article where it said this should not be done).
I now have 1. problems loggin in:
(Picture on top)
When I reload the page, it works and I get to the “Projects” page.

But now I am 2. not able to create a device profile, as it returns this error (I have tried many times):
(Picture in the middle)

My configuration looks like this:
(Picture on the bottom)


Thanks for helping!

Hi, @Mathew

I am Andres, a member of the Developer Support team.

I am sorry to say that once you delete your account there is no way back. The only way is to create a new account and start all over again.

Do you use any framework for the ESP devices?
In case you are using MQTT there are already some templates made by @TAustin . But unfortunately, the device types I need are not supported by the existing Edge driver.

What do you need? Additional device types can be added!

My profiles are a little bit troublesome.
I started to play with IoT controllers and decided to put them in my dishwasher and coffee maker.
The final goal is to have them in Alexa and Google Home.

The road is bumpy:

  1. there are no coffee maker capabilities in the ST, even though there is an icon for it. The closest I found is a cooker.

  2. there are two capabilities for dishwashers, but they don’t work as expected.
    2.1 The set of washing modes is hardcoded and the naming is not consistent with Samsung manuals.
    2.2 The available actions presentation is “wash”, “pause”, and “stop”. But in the routines there are “paused”, “drying” and “stopped”

  3. Neither of these device profiles is supported in digital assistants. They are displayed as unknown devices.

Could we maybe build a custom capability? What you need in a capability in order to be able to control from Alexa, for example, is a switch for commands, and something like a contact for conditional tests.

So far it takes too much of time. There are many small things that can go wrong with the custom capability, as there is no good validation on Json. And the examples are not there or don’t work well.
As for voice assistants I have a hope in the Matter, but likely will end up with some custom handlers.