Question on how to manage multiple garage sensors?

[SteveKlos]

Folks,

OK - our place has 4 different garage sensors.  I would like to setup a "goodnight" program.  This program would need to include if/then validations for each of the sensors.  I believe this should be straight-forward - basically I could have a "goodnight program" that does the following:

call close_gate program
call close_garage1 program
call close_garage2 program
 

Each of these programs will essentially be something like:
 

close_gate
      if sensor is open
            then turn gate on //this closes the gate

close_garage1
      if sensor is open
            then turn garage1 on //this closes the garage

close_garage2
      if sensor is open
            then turn garage2 on //this closes the garage
 

The question I have will this work without interruption?  Depending on the processing done on the ISY, it may only allow one program to run at a time - meaning the goodnight and possibly close_gate would run and that's it.  It's also possible that multiple threads can be operating at the same time meaning that all 3 close_ programs are likely to run at nearly the same time.

I just want to ensure I'm not running down a dead end alley here...

How do others deal with multiple garage sensors and a single program that needs to utilize the if/then capabilities of the ISY.

Cheers,

SK

[larryllix]

You may encounter people here telling you not to close a garage door without visual confirmation there isn't a toy, or a young child under the door.

I don't control my garage doors. I do notify the humans if they are open though.

While any door is open I continuously flash one lamp in the corner of my Gathering Room, with red flashing. If the Gathering Room lights are off and the bedroom lamp is turned on (automatically) I flash that lamp at full brilliance. I normally run that 100W equiv. lamp  about 10% at nights.
The bedroom lamp is flashed by an ISY program loop, while the Gathering Room lamp is flashed by a self-contained bulb code, so ISY doesn't have to do it.

Programs can just go down a list of items to fix. There is no need to detect whether they are on or off,  just set them to the position you want. With garage doors that is usually different, as the control toggles them open/close but there should be programs already in place that you would just call as subroutines, and those programs would decide what to do.

So it could look something like this.

If
     bedtime detected
Then
     set lamp1 to Off
     Wait 2 seconds
     set lamp2 to Off
    Wait 2 seconds
     Run GDO1 close program (if)
    Wait 15 seconds
    Run GDO2 close program (if)
    Wait 15 seconds
    $sSayGoodNight = 1  <---- trigger alexa routine
    Wait 5 seconds
    $sSayGoodNight = 0
Else
    -----

[jec6613]

1 hour ago, larryllix said:

You may encounter people here telling you not to close a garage door without visual confirmation there isn't a toy, or a young child under the door.

I don't have children, I have labrador retreivers, and multiple sensors on the opener that will bounce the door back, plus a camera.  Since there's no way we can see the garage door from inside the house, and no way without walking outside (even though it's a short distance) to close said door otherwise, I use an IOLinc and some keypad buttons.  That said, when the existing IOLinc goes, I'll be upgrading to something with even more logic and sensors.  

As far as running, it sounds like it would run just fine, but it's a bit kludgy.  I have my goodnight program set a variable, then use that variable to trigger other programs.  Ditto home, away, etc.

By doing this, it also lets me change out how the house responds in the middle of the night vs after sundown, including ramp rates and on levels for lights.

[larryllix]

3 hours ago, jec6613 said:

I don't have children, I have labrador retreivers, and multiple sensors on the opener that will bounce the door back, plus a camera.  Since there's no way we can see the garage door from inside the house, and no way without walking outside (even though it's a short distance) to close said door otherwise, I use an IOLinc and some keypad buttons.  That said, when the existing IOLinc goes, I'll be upgrading to something with even more logic and sensors.  

As far as running, it sounds like it would run just fine, but it's a bit kludgy.  I have my goodnight program set a variable, then use that variable to trigger other programs.  Ditto home, away, etc.

By doing this, it also lets me change out how the house responds in the middle of the night vs after sundown, including ramp rates and on levels for lights.

I use a lot of variable for lighting moods. Increasing the variable by one can brighten the lighting scenes (not Insteon, but combinations of Insteon, WiFi bulbs and strips with colours). When "borrowing" a lamp for flashing or other temporary uses, it is easy to restore the lighting in effect before the "Borrow" by replacing the variable value.

 

It's not usually your children. It would be the neighbour's baby carriage pushed by the older sibling under your garage door. The existing safety mechanisms in GDO could still break a child's. neck or back. They are not designed for that kind of protection. Paying law suit would not help you live in the neighbourhood, or sleep, if something like that happened.

[oberkc]

Regarding the programs, I have never knowingly run into a problem that I have associated with running simultaneous programs.  On the other hand, I have come to suspect that the signal collision avoidance routines built into insteon are not as effective as I would have hoped.  So...while the programs may run fine, if you have a lot of insteon commands blasting out at a given time, it seems that some can get lost or missed.  Wait periods in programs can help with this problem, but they ca also introduce some unexpected problems if the program condition changes during a wait.

[larryllix]

35 minutes ago, oberkc said:

Regarding the programs, I have never knowingly run into a problem that I have associated with running simultaneous programs.  On the other hand, I have come to suspect that the signal collision avoidance routines built into insteon are not as effective as I would have hoped.  So...while the programs may run fine, if you have a lot of insteon commands blasting out at a given time, it seems that some can get lost or missed.  Wait periods in programs can help with this problem, but they ca also introduce some unexpected problems if the program condition changes during a wait.

I think the PLM and devices have limited stacking facilities to create a polite transmission order but with a lot of Insteon traffic, that feature can potentially make it worse for battery operated devices that cannot detect powerline collisions. It seems Insteon battery devices just do a dumb algorithm by just banging the RF signal out a few times at random. This could make things even worse yet.

[apostolakisl]

The PLM, I believe, has a queue (either that or ISY does).  But either way, there is about a 1/2 second delay between multiple Insteon commands issued at the same time via ISY.

[larryllix]

38 minutes ago, apostolakisl said:

The PLM, I believe, has a queue (either that or ISY does).  But either way, there is about a 1/2 second delay between multiple Insteon commands issued at the same time via ISY.

Insteon protocol is supposed to include an anti-colision algorithm mechanism in every Insteon device. Either way enough packet stacking would likely cause problems for battery devices, that are blind to it.  I think.

[lilyoyo1]

My goodnight programs simply issues close and lock commands. I don't worry about the checking status of something. That way if the status is wrong the action still happens

[apostolakisl]

2 minutes ago, lilyoyo1 said:

My goodnight programs simply issues close and lock commands. I don't worry about the checking status of something. That way if the status is wrong the action still happens

Most garage door openers (all as far as I know) simply issue a toggle command.  So you can't do that.  I have my garage doors connected to Elk zones, so the status is never wrong.

[lilyoyo1]

6 minutes ago, apostolakisl said:

Most garage door openers (all as far as I know) simply issue a toggle command.  So you can't do that.  I have my garage doors connected to Elk zones, so the status is never wrong.

Setting the iolinc to Momentary C will allow this

[apostolakisl]

1 hour ago, lilyoyo1 said:

Setting the iolinc to Momentary C will allow this

I don't understand.  Momentary means the contact closure holds for whatever you set, say 1 second.  That is still a toggle.  It is ISY "pushing the gdo button".  Just like if I push and hold the button with my finger for 1 second.  If it is currently closed, it opens, if open, it closes.  Unless you have some different kind of gdo than I am aware of.  I suspect that the latest and greatest that have wifi and apps probably have an api that allows for separate open/close network commands, but I don't think many of those are out in the wild, plus, I'm don't even know that they actually do that.

[jec6613]

9 hours ago, larryllix said:

It's not usually your children. It would be the neighbour's baby carriage pushed by the older sibling under your garage door. The existing safety mechanisms in GDO could still break a child's. neck or back. They are not designed for that kind of protection. Paying law suit would not help you live in the neighbourhood, or sleep, if something like that happened.

Well, for one thing, I have multiple beam sensors crossing the path of the door from a few inches up to 3' off the ground, any one of which will cause an instant reverse.  Also, at least on my opener, the sensitivity on the back pressure sensor is as high as it can go - I can stop it with one finger, and have tested this.

And living in the sticks on some property takes care of most of the rest, and the cameras the rest, since I have much better visibility from in my house with the cameras than I do inside of the garage using the hardwire button.  I could meet the buzzer recommendations as well (not code required around here), and at some point will put that in, but since I can see the interior and exterior of the garage it hasn't been a high priority.  See also: the dual Labrador alarm system that would definitely alert me if someone was on the property.

[TrojanHorse]

Well, for one thing, I have multiple beam sensors crossing the path of the door from a few inches up to 3' off the ground, any one of which will cause an instant reverse.  Also, at least on my opener, the sensitivity on the back pressure sensor is as high as it can go - I can stop it with one finger, and have tested this.

And living in the sticks on some property takes care of most of the rest, and the cameras the rest, since I have much better visibility from in my house with the cameras than I do inside of the garage using the hardwire button.  I could meet the buzzer recommendations as well (not code required around here), and at some point will put that in, but since I can see the interior and exterior of the garage it hasn't been a high priority.  See also: the dual Labrador alarm system that would definitely alert me if someone was on the property.

Sounds like you’ve got this set up better than most. Given that newer GDOs allow app control the AHJs must believe that there is a safe way to do this. I think @LarryLix wanted to point out generally and sort of make a PSA that this isn’t always wise and I agree with him.

In the rare instances that I do this I do it through Mobilinc with the program associated with a camera so I can see as I do it.


Sent from my iPhone using Tapatalk

[larryllix]

44 minutes ago, TrojanHorse said:

Sounds like you’ve got this set up better than most. Given that newer GDOs allow app control the AHJs must believe that there is a safe way to do this. I think @LarryLix wanted to point out generally and sort of make a PSA that this isn’t always wise and I agree with him.

In the rare instances that I do this I do it through Mobilinc with the program associated with a camera so I can see as I do it.


Sent from my iPhone using Tapatalk

All my cameras have a few seconds of delay. If I saw a child entering under the door, it would a few seconds later and then the remote control command would likely take a few seconds more to stop the door travel. That doesn't even account for the human delay to find the dongle or click spot..

[TrojanHorse]

All my cameras have a few seconds of delay. If I saw a child entering under the door, it would a few seconds later and then the remote control command would likely take a few seconds more to stop the door travel. That doesn't even account for the human delay to find the dongle or click spot..

I agree with you. I rarely use mine (maybe once a year) and then it’s usually to open it for someone when we’re away and then they close it.

 

With our gates and garage and kids, my wife and I call if one of us is out to make sure it’s safe to open. I’m concerned even about a toddler finger in the garage door track when someone opens the garage door. Even the standard remotes can be dangerous because a garage door is dangerous.

 

I suppose a neighbor kid could do a wheelie in the stroller and park it under the door while someone is leaving and they simply using their visor button. There’s no substitute for a human eye.

 

 

Sent from my iPhone using Tapatalk

[jec6613]

10 hours ago, larryllix said:

All my cameras have a few seconds of delay. If I saw a child entering under the door, it would a few seconds later and then the remote control command would likely take a few seconds more to stop the door travel. That doesn't even account for the human delay to find the dongle or click spot..

Mine is around 5 frames of delay at 15 FPS, so 1/3 of a second, on my LAN.  From Joburg it was about 2-3 seconds transiting the undersea cables.  But if a child enters under the door path, it stops and retracts due to the curtain sensors anyway, and they'd also have to come from somewhere, which is all camera covered.  If a child can run 150m from out of sight to the door in less than 12 seconds when the world record is 9.58 to 100m, then I'm calling Guinness.

[larryllix]

1 hour ago, jec6613 said:

Mine is around 5 frames of delay at 15 FPS, so 1/3 of a second, on my LAN.  From Joburg it was about 2-3 seconds transiting the undersea cables.  But if a child enters under the door path, it stops and retracts due to the curtain sensors anyway, and they'd also have to come from somewhere, which is all camera covered.  If a child can run 150m from out of sight to the door in less than 12 seconds when the world record is 9.58 to 100m, then I'm calling Guinness.

Not too much of that Guinness!

I was looking for a multiple beam sensor originally. Do you have any links to one of those?

[apostolakisl]

I close mine remotely if I am going to bed and forgot it open.  Not a lot of toddlers in my driveway at midnight.  Actually, not a lot of toddlers in my driveway ever.  At least not since my son stopped being a toddler 5 years ago.  I live on 7 acres and there just aren't any kids wondering around anywhere at any time on my street.  I see more bobcats on my property than toddlers.  And if you know anything about bobcats, you'll understand what I mean.  I have trapped birds in the garage, however.

[larryllix]

1 hour ago, apostolakisl said:

I close mine remotely if I am going to bed and forgot it open.  Not a lot of toddlers in my driveway at midnight.  Actually, not a lot of toddlers in my driveway ever.  At least not since my son stopped being a toddler 5 years ago.  I live on 7 acres and there just aren't any kids wondering around anywhere at any time on my street.  I see more bobcats on my property than toddlers.  And if you know anything about bobcats, you'll understand what I mean.  I have trapped birds in the garage, however.

The one that bothered me was: now having a hatchback style opening back door on a CR-V,  I have seen several rear vehicle doors at body shops, last time I was there. They become a very expensive repair into several thousands of dollars to replace. I have several laser pointer type parking guides on the garage ceiling, and the first spot that falls off the dash gives a tested clearance, including the hinge hardware, clear of the vehicle by a few inches, while the door is open and during any part of the opening arc..

Just another consideration for blindly closing garage doors. OTOH: That could also happen while you are closing it from the wall PB,.

[jec6613]

6 hours ago, larryllix said:

Not too much of that Guinness!

I was looking for a multiple beam sensor originally. Do you have any links to one of those?

Mine came with the house.  When we had the door and track replaced by UPS (who damaged it) is when I got all of the in-built sensors calibrated and dialed in.  Also since we now have a lightweight door and previously had solid wood, we have a massively overpowered GDO, which really helps with the auto reverse sensor.

I believe it's a seco-larm curtain sensor, similar to this: http://www.seco-larm.com/sensors/curtain-sensors/E-9622-4B25

[apostolakisl]

6 hours ago, larryllix said:

Not too much of that Guinness!

I was looking for a multiple beam sensor originally. Do you have any links to one of those?

we do have a tahoe and on occasion it isn't parked in far enough, but when parked as it is normally, the door opens without hitting the garage door.  But yeah, that is a risk.  Though I would expect damage from the GD to be quite a bit shy of destroying the back door.  Paint scratches, maybe break the windshield wiper, but not structural damage.  It should just push the door into a more shut position.

[larryllix]

20 minutes ago, apostolakisl said:

we do have a tahoe and on occasion it isn't parked in far enough, but when parked as it is normally, the door opens without hitting the garage door.  But yeah, that is a risk.  Though I would expect damage from the GD to be quite a bit shy of destroying the back door.  Paint scratches, maybe break the windshield wiper, but not structural damage.  It should just push the door into a more shut position.

I bent a side door by backing into the GDO track. I bent the track back with my hands and it's been fine, not showing any dents for last few years. It was minor. The Rav4 door had the corner metal rolled over but bent it structurally, slightly. That one cost me almost $3K to replace the door with a junkyard, lower model (no coatings) door. Mind you the bodyshop people all told me the window tracks are welded in and cannot be made to not bind, on a scrollable door window. Then there is all the power gadget switches.

While I was at the bodyshops I saw several bans with the awning type doors with dents in them. I was told by different shops the door were write-offs, as the painting of half the vehicle was the larger part of the costing, anyway. Insurance coverage (big deductibles) was part of the solutions, I am sure.

[larryllix]

40 minutes ago, jec6613 said:

Mine came with the house.  When we had the door and track replaced by UPS (who damaged it) is when I got all of the in-built sensors calibrated and dialed in.  Also since we now have a lightweight door and previously had solid wood, we have a massively overpowered GDO, which really helps with the auto reverse sensor.

I believe it's a seco-larm curtain sensor, similar to this: http://www.seco-larm.com/sensors/curtain-sensors/E-9622-4B25

Thanks! That doesn't look easy to interface to a GDO and I can't see any pricing.  Actually, my original drive was having a side opening back door on a RAV4 and the garage door coming down on the edge would have demolished that door. It becomes very protruding when they open sideways. Other than that, completely useless format.

I raised my GDO beams up to the level of the lower tip of the vehicle door in an attempt to avoid that accident. I realised it helped endanger the very child problem we have been discussing. However I live in the rural and have no small children in the area. Worst is coyotes, skunks, and cats, but I would be proud to squish one of them.

[apostolakisl]

57 minutes ago, larryllix said:

Thanks! That doesn't look easy to interface to a GDO and I can't see any pricing.  Actually, my original drive was having a side opening back door on a RAV4 and the garage door coming down on the edge would have demolished that door. It becomes very protruding when they open sideways. Other than that, completely useless format.

I raised my GDO beams up to the level of the lower tip of the vehicle door in an attempt to avoid that accident. I realised it helped endanger the very child problem we have been discussing. However I live in the rural and have no small children in the area. Worst is coyotes, skunks, and cats, but I would be proud to squish one of them.

seems like you could set up some mirrors at various angles and reflect the beam back and forth up the track.  Challenge would be keeping them from getting out of whack.