Slow Rules?


#1

I have quite a few simple lighting rules setup for motion > light. They are kinda slow, slower than I expected for a completely local setup.. roughly 2-5 seconds. My motion sensors flash a little light when they detect motion, so I can see that they activate immediately. So where should I be looking for slowdown issues?

Additionally, I have one room that's using RM instead of SL. It triggers a virtual switch on motion. A secondary rule turns on lights and the TV when that switch is on. (this is all so that the lights can be dimmed without the motion setting off 100% again) This whole setup is very slow... on the order of 10-15 seconds. Should this be a triggered rule?


#2

Break it down to it's components to establish a baseline for that parts of the rule. Substitute your motion sensor for a virtual motion sensor and manually trigger is in its device details. How fast does the light turn on?

If slow, trigger the light directly from the driver. How fast it is?

If fast, trigger the motion sensor and watch the status in its driver details to see how quickly it changes. The light blinking on the motion sensor just says that the hardware detected motion, but it's not an indication that the device's radio successfully sent the message to the hub in a fraction of a second.

It's possible the rule that is turning the TV on is having to wait for that device to send back status too. Remove that if everything above checks out to be fast, and retest. Something is causing a delay and it's not motion lighting.


#3

Also, use live logging to see if the motion sensor is actually reporting when you see the light flash.


#4

In addition to the great tips above...

What type of motion sensor? What type of lights? Zigbee, Z-Wave, WiFi, etc..? Brands and models can help to identify known problematic devices.

Would also be good to share screenshots of your rules.


#5

Probably good to know other custom applications and drivers you may have installed. I experienced a similar situation a while back, the MLAs were really slow as well as other applications. With some custom application clean up, things have improved significantly.


#6

I'll do those tests when I get home from work for sure. Also.. taking screenshots, I had forgotten I edited that theater rule to use a private boolean instead of virtual switch. I also apparently tried it as a triggered rule. TBH, I don't know that I fully understand the difference between a normal rule and a triggered rule.

All rooms use the GoControl motion/temp sensor (z-wave) and Lutron dimmers. A couple do Lutron and handful of Hue lights.


#7

I've been pretty sparse on custom apps and drivers I think.

Apps
-Keenect Lite
-Cobra Average All
-MyQ Lite
-Rheem Econet
-NST Manager

Drivers (In addition to the required child devices for the apps above)
-AeotecMultiSensor6 (cSteele's advanced driver)
-LeakSmart Valve
-Logitech Harmony Hub Parent
-Ring Spotlight
-Z-Wave Lock (I don't think this one's in use.. I had to use it to configure my Schlage lever lock)


#8

I wouldn't call that sparse... Mainly because a number of them are LAN based apps/drivers, and latency on LAN connections (if not handled excellently in the code) can cause slowdown issues.

I'm not saying there are any issues with those specific apps/drivers, just pointing out user LAN based apps/drivers are more of a concern than user zigbee/zwave drivers.


#9

To be a little blunt here.. this is why I've shied away from some of the interesting apps I've seen. The go-to around here is to quickly blame custom apps/drivers for issues... so why bother using them if they either cause this many issues and earned all the blame and/or get off-the-cuff blamed for support issues regardless? The apps I've installed are the ones I need to connect certain devices. I would remove them if I could.


#10

No argument... Just stating opinions and anecdotal findings from the past.

There have been multiple times where poorly made LAN apps/drivers did cause hub issues, that's just a fact. So it isn't unreasonable to point out that those could be an issue in some scenarios.

With so few of them, and you having repeatable/testable slow performance, it shouldn't be hard to disable them one at a time and see if your response time changes. If it does, then it is one of those. If it doesn't then it isn't.

Before I would do that, though, as mentioned by others above, I would look at the logs in detail when you trigger motion and make 100% sure the delay isn't on the device side in reporting the data to Hubitat. If it is, then not much you can do about that other than replace the motion sensor with something faster...

I would specifically look at the time between the incoming motion event, the rule trigger, and the output event (may need debug logging turned on on the devices).


#11

No I understand.. just stating why I've actively tried not to use custom stuff.


#12

It took me a bit to wrap my head around this, too.

Here’s how I understand it, and I could be mistaken:

Example: Let’s say you have a rule where if a motion sensor goes to ‘inactive’ then a light turns off.

Normal rule:
When motion sensor ‘inactive’ is true, turn off light.

  • any time the state of that sensor changes (active or inactive), the rule runs to check if it is true or false
  • thus the rule is listening and runs a check on the ‘truth’ action whenever the motion sensor is active or becomes inactive
  • you can set a ‘cancel on truth change’ which is useful if you want the lights to turn off 5 minutes after the motion stops unless the motion starts again, in which case the ‘off’ action is canceled. Until the motion becomes inactive again, at which point the 5 minute timer starts again before turning the light off.

Triggered rule:

  • The rule sits dormant until a trigger event happens.
  • once the rule is triggered by the specific event, there is a one-time check of the rule to determine if it is ‘true or false’ and then runs those actions accordingly
  • a triggered rule can’t have the ‘cancel on truth change’ as the rule is not constantly listening to see if the truth has changed - it’s only a one time check after the trigger event.

Regarding motion sensors and zigbee vs z-wave:

  • personal anecdote as I have a few zigbee motion sensors (sylvania/osram) and zwave motion sensors (Aeotec).
  • my zigbee sensors trigger actions usually in 1 second or less.
  • my zwave sensors trigger actions in 2-3 seconds. There is a noticeably longer lag than with zigbee for whatever reason. I don’t know if it’s the sensor, the protocol, the hub, aliens, whatever. :slight_smile:
  • I am not at all saying this is the root cause of your issue, just sharing a personal anecdote from the behavior I’ve seen on my setup

#13

These sensors are notoriously slow to transmit motion events. I typically was halfway across a room before the lights would turn on using these devices. I now use Lowes Iris V2 Motion/Temp sensors which are nearly instantaneous. The difference is amazing.

During last week's Hubitat Live YouTube session, @mike.maxwell went over a broad selection of motion sensor that are compatible with Hubitat. If I recall correctly, those GoControl sensors were near the bottom of his list (they may have even gone in the trash can! :wink: )

Here is a link to the video, at the exact time when Mike discusses the SlowControl Motion sensors...

Update: While Lowes Iris is out of business, you can still grab sensors on ebay at a great price.


#14

I'll watch the video.. If I'm going to switch sensors, I'd like some that have more than just Motion/temp. I got an these go control ones when I was still on Wink.


#15

I understand the desire for a multi-sensor... Unfortunately there just aren't too many of them available that are 1) reasonably priced, and 2) work well.

~$6 per sensor is really a great price for the Iris v2 zigbee motion sensors. Great battery life and performance. They are also very small.


#16

Or the Nyce ones are pretty badass with a 5 year battery life. BUT closer to $50/ea.


#17

What else are you wanting from a "motion sensor"? The general multi-sensors out there are not very good. They all do some things well and the rest of the multi-pattern poorly....

Zigbee sensors are faster than z-wave. Period.

Best sensors... my opinion....

  1. NYCE motion sensors (temp/humidity included) fast response all functions work well.
  2. Hue motion sensor. Only one I currently know of with active Lux reading. Meaning the lux updates in realtime and can be actually useful for lighting control. [EDIT: include temp that works as well]

I've test a lot of motion sensors and those currently are my top 2 picks. I have the zooz and 4-in-1 and the rest and they are "ok" but nothing spectacular. The Zooz ZSE02 is the fastest of the z-wave bunch and is only motion but it's the size of an echo dot! The other eye-ball sensors zooz zse09, dome, neo etc etc all clones of the same thing and all have false positive behaviors of being overly sensitive.

The Fibaro is a great sensor but is low on the list only because of the price. It has good performance but it is still slower than the NYCE because of being z-wave.


#18

I have both the Hue and Iris and while the Hue work well my Iris sensors have much better battery life and seem to have a better FOV. I tried mounting Hue's up in a corner of the room and it would barely detect anything, they seem to only work well at chest level.


#19

I have 5 different types of motion sensors, and they all have pluses and minuses.

  1. Hue motion sensor - Zigbee. I had problems getting it to work with my SmartThings hub, so I moved it back to my Hue hub. I've left it there, so I can't say how it works with HE. But with Hue it's been great. It uses lux to determine if the light should be turned on or not (in a Hue routine), and is right most of the time.

  2. SmartThings motion sensor 2016 version - Zigbee. Hard to pair, but once there works OK. Reports motion and temperature. One of mine just would not re-pair correctly, no matter what I did. I retired it. Reports temperature and motion. (this version is still available)

  3. SmartThings motion sensor 2018 version - Zigbee. Pairs easily. Better range than the 2016 version. Has been very reliable. Reports temperature and motion. Relatively inexpensive.

  4. Fibaro motion sensor - Z-wave. Paired OK. Reports motion, temperature, vibration, and lux. It seems to be pretty responsive as far as motion, temp, and vibration go. Lux is so-so. Expensive.

  5. Dome motion sensor - Z-wave. Paired easily. Reports motion and lux. It has a wide motion range. I'm not sure how it does on distance. I'm using it only for the lux, and that seems to work pretty well. It's not as expensive as the Fibaro, but it costs more than the ST.


#20

So then, do you use Rules or triggered rule to control your lights?
I also have very long delays in motion lights working. I used all triggered rules for my lights.
BTW, my motions are all hardwired and is a zone on my security panel that talks to my hub via a driver.