I am new to HA, however I have picked up that some of the functionalities asked for here is partially tended to, with templates.
I come from Smartthings where typical the automation for a device is offered in a mini smartapp.
My suggestion is for a range of integrations to manage simple automations for devices, eg.; motion sensor to bulb, switch to blinds.
This would be easier to set up in the GUI and enable a full sequence flow to be handled in one automation.
What’s difficult to setup in the GUI? Templates are mostly used for more conplex automations.
Did you create some automations already or is this all theoretical?
Most proprietary products such as SmartThings is like riding the bus where software such as Home Assistant and the like is more like owning a car. They both essentially go from point a to point b but one is entirely configurable and complex as well as offer pit stops at point a.1, a.2, a.3 and so forth. Because of this there is no way to cover every aspect that some would think would be more simple so instead you are given a basic car and presented a parts store that contains everything you could possibly need and if if it doesn’t already exist it is possible to create.
1 Like
I did set up automations.
Similar to what I had previously; motion sensors, lights, blinds, vibration sensor, water leak sensor with and without conditions.
It just struck me it could be done leaner, more intuitively and with less automations - something as simple as a motion sensor to light pairing requires to different automations.
In comparison Smartthings handles the full flow in one automation “package”.
I do realise the difference between the two concepts. I considered that when switching, and I am still glad I did. For the same general reasons you argue for in the bus/car comparison.
However, that general comment can hardly be a decision to not make things easier or more lean in HA if possible?
It would not take much to have a motionsensors to lights pair be handled in one simple flow rather than multitude of individual automations.
Can you show an example of your motion light automations? I think they are pretty straightforward. Especially with the “new” device automations where it suggests what you want to do when for example motion is detected.
They can be handled in a single automation. I guess I don’t understand what you are meaning.
Theres is an example of the controls for a light with a motion sensor:
- id: ‘1599340160329’
alias: Vaskerum
description: Tænd
trigger:- device_id: c4931f9eeeaf443c850d2abe87d2f0ae
domain: binary_sensor
entity_id: binary_sensor.vaskerum
platform: device
type: motion - device_id: c4931f9eeeaf443c850d2abe87d2f0ae
domain: binary_sensor
entity_id: binary_sensor.vaskerum
for:
hours: 0
minutes: 10
seconds: 0
platform: device
type: no_motion
condition: []
action: - brightness_pct: 100
device_id: 44e1d0b1dd0e4c758ad3555880880f00
domain: light
entity_id: light.vaskerum
flash: long
type: turn_on - device_id: 44e1d0b1dd0e4c758ad3555880880f00
domain: light
entity_id: light.vaskerum
type: turn_off
mode: single
- device_id: c4931f9eeeaf443c850d2abe87d2f0ae
Reading the link provided, is also for setting up one action only?
Then I may just need to be educated.
Lets for the sake of the argument say that I want to setup a simple light that turns on at motion from a sensor and turn it off again after x seconds if no motion or presence is detected.
I could only do that by setting up two automations.
Incorrect, that could be done with a single automation. You should read up on conditions. Even if you had a more complex situation with the use of sensors and binary sensors and helpers(bools) and everything else you could still achieve a single automation with more complex variables. Not to mention even if you didn’t like how yaml is used you could still use node red or appdeamon to handle all of that in a completely different matter.
Have you really thought this thru?
What you are basically saying is that you want the devs to create automations for you?
or if you aren’t asking for that then I can’t see how Smartthings would be any easier than using the UI editor to set up two very simple automations.
I’m sure that in Smartthings you still had to go thru some kind of UI to select which motion sensor turned on which light. And then you would have had to select that the motion sensor turned off the light after a certain period of time after motion stopped.
You can already do all of that very easily in the editor as it is right now. There’s no reason for any “multitude of automations”. It’s only two.
I don’t see how Smartthings could be that much easier for that particular use case?
1 Like
I cannot see how you get the impression that I want custom automations created for me.
Try to forget that I ever mentioned coming from Smartthings. We probably all come from something simpler and less bespoke. It was in an effort to refer to a live example to increase understanding.
It was not a request to “dumb it down for me” or for others to imply how clever or matured you are in using HA.
I am merely suggesting that the current GUI for setting up individual automations are developed to handle handle a sequence of automations. This will enable creation of flows that are typically wanted from a device.
The learning curve of HA is kind ever changing to a degree. They have made vast improvements to the UI just within the year I have been using it and leaps and bounds from what I gather to those who been using it prior. But once you start to really poke around and understand what does what and get a solid grasp with how everything comes together you will quickly understand why it is the way it is and even with that, they in fact are continuously improving it so even after you are elbow deep like many of us they will likely improve upon it to better touch what you are trying to get at.
If anything, the ability to make the automations are pretty simple but sometimes the explanations and examples of what does what is likely the culprit for many.
Ok. If that is a feasible way to handle this.
The first I ever read on conditions was this: https://www.home-assistant.io/docs/scripts/conditions/
“Conditions can be used within a script or automation to prevent further execution. When a condition does not return true, the script or automation stops executing. A condition will look at the system at that moment. For example, a condition can test if a switch is currently turned on or off.”
So, my understanding is that a condition make an automation not execute. Not stop it after the execution and revert the device back to its previous state.
What you a saying (based on the sensor / light example) is that I can have an automation to turn on the lights AND with the right condition setting, make light turn off again (based on the condition parameters)?
You have Triggers, Conditions and Actions. You can also have conditions IN actions. The initial main Conditions can stop or allow an automation to execute at all while conditions in the Actions can be used to stop progression of an automation beyond that point. There is even a new thing called Choose which I assume is more like If this Than that but don’t quote me on that since I haven’t used or even read up on it. Inside the actions is where you can put pretty complex scenarios together to achieve the multiple things you want a single automation to tackle. I would take an hour to read up on every possible option you see listed in the UI to get that initial grasp of all the tools presented to you and what they do.
Concur. That is close to what I call an integration to define a sequence of typical automations.
Must admit i searched for “WTH” topics only when I skimmed for possible dublets.
You can use choose, but you’ll have to move away from the device oriented triggers. It’s still a bit complicated because you’ll need to peel the trigger information through a template. Either way, you can use choose with the triggering entity to go one path or another.
THX that is something I will look into.
For the long term, I will support the Why the heck don’t we have prebuilt / shareable automations?