Securifi Community Forum

Securifi Products => Almond+ => Topic started by: penright on April 12, 2017, 07:58:15 am

Title: Rule behavior not as expected with Linear GD-004
Post by: penright on April 12, 2017, 07:58:15 am
I have a rule that when the overhead garage door (OGD) is opening and a light switch is on, then turn on some lights for 10 minutes.
I am  using the light switch for my "between sunset and sunrise" workaround. Look at this link for details https://forum.securifi.com/index.php/topic,5131.0.html (https://forum.securifi.com/index.php/topic,5131.0.html).

If I use the Almond app to open the OGD then the rule fires as expected. If I use the OGD wall switch it does not fire, but if I am looking at the app it is showing the garage door opening.
Is this expected?
Title: Re: Rule behavior not as expected
Post by: fillibar on April 12, 2017, 09:21:04 am
How is the OGD monitored, with a sensor alone or is this part of a garage door controller unit?
My thought is that if it is a unit maybe it reports the door opened by the switch differently?
Title: Re: Rule behavior not as expected
Post by: penright on April 12, 2017, 12:51:14 pm
How is the OGD monitored, with a sensor alone or is this part of a garage door controller unit?
My thought is that if it is a unit maybe it reports the door opened by the switch differently?
Sorry, should have mention it is a Gocontrol.
As I said, the app was showing the door opening, so I am assuming the A+ had to have seen a report from the sensor that the door was opening.
Since the app reported the door opening no matter which method the door was opened by (wall switch or app), but the rule only fired when the door was opened by the app, therefore the A+ was in a command mode, rather than a read mode, I bet there is a bug. I bet the rule is only check when the A+ is commanding, not just changing states. Just a guess :-)
Title: Re: Rule behavior not as expected
Post by: fillibar on April 12, 2017, 04:13:38 pm
If the A+ gets the state change it should trigger from it. I checked the manual and what little is online about it's commands. It only gets the door state by the tilt sensor, so it only sends "one type" of open state to the Almond. It does not care what caused it.

So I do not see any immediate reason why it would not trigger.

Can you make a super simple rule to test it?
IF Garage Open THEN Light On

Results when garage opened by app.
Results when garage opened by button.
Results when garage opened using wireless garage door opener.

That might help narrow it down a bit.
Title: Re: Rule behavior not as expected
Post by: penright on April 13, 2017, 10:40:05 am
Results when garage opened by app.
Worked.

Results when garage opened by button.
Did not work. I did not try the wireless since the button did not work, figured it would be the same.


IF Garage Open THEN Light On
As I was responding, I noticed you said , Garage Open, I am using Garage Opening. Could that be the difference?

I don't have a public IP, so I can not connect from my office. If I send the logs from the app, can I get a copy somehow?

Title: Re: Rule behavior not as expected
Post by: fillibar on April 13, 2017, 11:10:32 am
I do not know of a way to get a copy of the logs sent via the app. Sorry.
I also do not know if there is a real difference between Open and Opening. I do not have a GD-004 (although I have looked at it and the Aeon Labs variety). I do have one of the simple tilt sensors on one of our garage doors (I have multiple of the same type but for some reason the others refuse to pair correctly). It says Opened and Closed.
The reason I mentioned the wireless was just to prove that it was the same as the button. If it was different (matched the app) then that could also be useful to figuring out the issue (maybe).
Title: Re: Rule behavior not as expected
Post by: penright on April 13, 2017, 01:50:39 pm
I also do not know if there is a real difference between Open and Opening.
Look like there is ...

The GD-004 tilt sensor is reported via the GD-004.
So I thought I would grab a screenshot so you could see my options. While there, I grabbed one of the device history and some of the magic was revealed ....

Here they are, the interesting one is the 'Recent Activity' ...
(http://i1351.photobucket.com/albums/p781/penright/StuffForForums/Screenshot_20170413-123035_zpsw8qve1hg.png) (http://s1351.photobucket.com/user/penright/media/StuffForForums/Screenshot_20170413-123035_zpsw8qve1hg.png.html)

(http://i1351.photobucket.com/albums/p781/penright/StuffForForums/Screenshot_20170413-123149_zpseszq1lsx.png) (http://s1351.photobucket.com/user/penright/media/StuffForForums/Screenshot_20170413-123149_zpseszq1lsx.png.html)

The 6:02am and 6:03am was me leaving for work (6:02 was button and 6:03 was remote). The 9:30am and 9:31am was my wife testing the button. Then 9:31am (2nd one) was the app commanding it. Then the 9:32am was here pressing the button to close.
What is interesting is difference between button and app activated. Now I am not sure how to fix this ....
Looks like the GD-004 distinguish between activations methods.

Can you have "something and (something or something) and something" ?

Title: Re: Rule behavior not as expected
Post by: fillibar on April 13, 2017, 02:57:11 pm
The solution would be two sets of Rules to cover both scenarios.
It is interesting to me that it reports it both ways. You would think they would end up being the same.
It also never reports closing.

Seems like some odd behavior from the GD-004. I wonder if anyone else that owns it has hit this. I am going to change the thread title to mention GD-004 so hopefully it draws other current users.
Title: Re: Rule behavior not as expected
Post by: penright on April 13, 2017, 03:26:33 pm
It also never reports closing.
It does, it was not in my snapshot.

The solution would be two sets of Rules to cover both scenarios.
Ok, got it, will try it tonight. Looking at the log, I see my wife has left the building. :-)

It is interesting to me that it reports it both ways. You would think they would end up being the same.
Yea, but it does explain everything. Actually the fact that it has one event for the button/remote vs another form the HA(GD-004) could come in handy under the right circumstances. Can't think of one right now, but ....

Anyway, I was expecting the "open" to keep the rule running as long as the door was open. That is why I was using "opening". I bet with the two rules, I can get what I need. :-)
Title: Re: Rule behavior not as expected with Linear GD-004
Post by: fillibar on April 13, 2017, 08:09:35 pm
Looking forward to hearing how it works. I am still looking for a deal on one of these or the Aeon Labs one... so it is always good to see how they work out for people.
Title: Re: Rule behavior not as expected with Linear GD-004
Post by: penright on April 14, 2017, 08:39:56 am
Looking forward to hearing how it works.
Well, it worked kind of ....
I had my grandson for the weekend and all to myself last night, so I did not get to play with it.
But this morning, the lights did come on. The weird thing is when I looked at the activity ....
You can see the door went 'opened' (button) and 'closed' (remote) at 6:04am and 6:05am respectively.
Both lights (SW1 & SW2) switched on at 6:04am
SW1 switched off at 6:09am (5 minutes) but SW2 was 6:14am 10 minutes. It was programed for 5 minutes just like SW1.
Couple of notes...., looks like my wife or granddaughter went into the garage and turn the light on last night at 6:28. Grandson is just two, so I don't think it was him and I don't remember going back in. :-)
That is why you see me flipping it off at 6:02am before the test. Also the screenshots of the rules may not be clear, but SW2 on is 0 sec and there is nothing after SW2 off.
Let me post my screenshots ....
(http://i1351.photobucket.com/albums/p781/penright/StuffForForums/Screenshot_20170414-065141_zpsbepq8dmw.png)
(http://i1351.photobucket.com/albums/p781/penright/StuffForForums/Screenshot_20170414-065113_zpsm5dmde6q.png)

(http://i1351.photobucket.com/albums/p781/penright/StuffForForums/aacadd0c-6dd6-4367-add0-b39e10d64c9f_zps5d0jo7yu.png)(http://i1351.photobucket.com/albums/p781/penright/StuffForForums/f08766ed-0f87-47d4-a199-e6c71e7b7f86_zpsx1lxvdrj.png)
Title: Re: Rule behavior not as expected with Linear GD-004
Post by: fillibar on April 14, 2017, 08:51:18 am
Switch 2 is taking 10 minutes because it has the 5min delay from switch 1, then it's own 5 minute delay. If you want them both off at the same time you would remove the delay from. Switch 2. All the actions happen in order, but without delays they happen so quick it seems they are all at the same time.