HE's biggest weakness :crossout: opportunity

Honestly I understand where you are coming from. Its kind of more than just something that is solved by a UPS and its not just power, but also internet outage (for cloud devices, we all have one) that it doesnt always recovery from.

Now I wouldn't say its an issue with HE per se but something that would be addressed at the driver/app level. In most of my apps I have them do things like clean up after a reboot or force a full refresh every now and then just in case to address things like this.

An example I had recently was with sharptools. Over night my internet went out. When I woke up and finally rebooted my modem sharp tools was all out of sync. No matter what I did I couldn't get it back in sync from the dashboards. But from the app itself I had to go into it and resave it so it forced a full sync and everything was back in business. I would expect that to be automatic.

Even with a zwave device. If it was on when the power went off and is not set up to restore state when the power comes on, it will be off and HE will still think its on. So I usually have HE polling those devices to see if they are alive and get an update in case they are out of sync. But that is in my driver itself.

So ya.. its a valid argument that it doesn't handle power failure well. Just need to find ways around it.

(And yes a UPS is critical for HE if you have frequent power outages)

If a device does not report, the hub has no specific awareness of that. An automation can be easily implemented that checks this (which would be completely reliant on the hub being powered up to work). But there is no reason for the hub itself to have this awareness absent an automation.

You are in control of your home automation system. The hub certainly doesn't know that a power failure occurred unless you have implemented a specific system to deal with this situation, nor could it have such an awareness otherwise. Without getting into a pointless debate about backup batteries, if power is pulled from the hub, it isn't a hub anymore. When power is restored, it reports the event of the system coming back up, but it has no awareness of its environment unless you have put those capabilities in place.

Ultimately, it is on you. You design your home automation system. You certainly shouldn't trust it with anything beyond its capabilities.

No, it's not astounding at all that problems such as you raise have not been solved. If this is part of your expectation set, you will continue to be disappointed.

9 Likes

With all due respect @bravnel - and I say that sincerely -
It's is precisely these issues that HE is in a perfect position to resolve. Efforts into these areas, even if they were exploratory would put HE into a position of strength as the other manufacturers like aeotech and samsung bring product into direct competition. At some point HE has to begin to take a leadership role. With the resources, experience and community support HE currently enjoys, now is a good time to roadmap toward the future.

1 Like

I already mentioned this in my previous posts but let me repeat my point again:
"Hub is Single Point of Failure, no doubts here".
Hub fails - nothing is working. And it does not really matter how good a recovery plan is.
Another point is:
" The Devices should be able to talk to each other directly bypassing hub".
This days there is no technical problem to achieve this and it is already
partially implemented by introducing the "direct association".
For instance, the entire Insteon system (about 20 years old) was built around "scenes".
Furthermore, each really smart device should/must have built-in mini RM.
Again, technically this is not a problem.

These issues are not resolvable in a hub in this price range or with this architecture. At this point, we don't see this as a meaningful opportunity for Hubitat. Thanks for your suggestion, but it is headed for the circular file.

6 Likes

I expected nothing less. that last comment made my FB wall for 800 of my closest friends to see.

Lutron RadioRA 2 has 2 functions that address these types of issues. I can see some value as part of the Hubitat system rather than as part of Rule Machine.

  1. Run events missed during power outage. Runs all scheduled events after outage.
  2. Catchup Events - when the mode this setting will catchup all the events for the active mode since midnight.

Is that supposed to be some sort of social media threat? Lol

I get that you might not like the answer, but he answered you honestly and respectfully.

10 Likes

No!!! I let me friends know I get kicked around in other places too!

Ah then never mind. I'm lost. It's not uncommon LOL

2 Likes

This could have disastrous effects in a Hubitat system, unless some steps are taken to specifically identify those actions that should or should not be taken. A Lutron system is dealing almost exclusively with lights, and turning a light on or off is unlikely to disrupt things. The automation challenge to identify which missed jobs to do is itself daunting, as there is no a priori way to know when an outage happens or for how long. Should a rule that had a delayed action pending still perform that action? Maybe, or maybe not.

I hear all of this and agree that "some" level of resilience is important to "many" folks. For some, what they do in home automation isn't that "critical". I have taken the steps necessary to make my setup as resilient as I need. Nothing is 100% fault tolerant and to make a home automation system even approach fault tolerance is an expensive venture. Auxiliary power, multiple diverse internet sources, battery backups on hubs AND repeaters, you could go on forever. But for a hub that costs $120 what do you expect it to do? High end systems like Control4 and others don't do much more after outages (and what they do you pay for). Remember that anything additional "built-in" to the hub cost money, the cost of which will be passed on to everybody that buys one (or two or three or four...). Most probably wouldn't need the extra, some would.

Hubitat has many tools that you can use to mitigate issues that may occur during a power failure. Remember that many of your endpoint devices are going to be dead unless you have a whole home generator that kicks on automatically with a whole house UPS to carry it over the transition.

For me, I have battery backups, with power monitoring on my hubs that set hub variables when the power fails. After a prolonged outage my hubs shutdown automatically. When the hubs restart they run a routine that checks the hub variable and performs some tasks that I deem appropriate (such as turn off the zigbee bulbs, etc). To be honest, they do what I (and probably most others) need.

Just my two cents....

4 Likes

Yes, of course there would be design issues with a more complex system (such as making the setting more granular than a global). Yet people are writing RM rules to do so. Are the benefits worth the cost to implement? That's for you to figure out.

Yes, this could be done even as we speak if Hubitat is powered from some sort
of UPS and UPS is plugged in power reporting outlet.
Instead of power reporting outlet you can use some sort of DIY power outage
detector. Use alway powered 120V (for the USA) relay and wire NO or NC relay
contact to the Contact Sensor. Or instead of relay use a small electromagnet
and attach Contac Sensor to it.

But if you are going to do all of this DIY on the hardware side, then why is it a problem to do the software side DIY (and make an app or RM rules to refresh devices after power up)?

Or am I missing the point again?

If DIY is OK, anyone can do that today - all the capabilities are in place. Get a power reporting plug, a Ring Extender that reports power status, etc and make RM rules or app to poll things for status.

3 Likes

In short - because I am EE.

Should this be all part of a kit supplied by Hubitat?
Maybe an optional kit that can be purchased from Hubitat.
120 has now gone up to~250?

HE staff. I see a revenue string here. :wink:

2 Likes

I hear ya... But the hub isn't going to be everything for everyone out of the box. At some point people have to decide to use the in-box capability to do it, or not. No one is going to come configure the hub for you.

I don't mean that snarky - my point is that there is no end to what it "could do" by end users versus what it is "worth doing" out of the box. Everyone has different needs, and the list of wants is never ending.

To be clear, I don't care one way or the other if they add it (or any other feature) - as long as it is optional. I for one wouldn't use it, and would turn it off in favor of my rules/apps I made that do exactly what I want.

Oh, and side note - you can't poll/refresh battery devices anyway unless they are a frequently listening device type like door locks usually are (but contact, motion sensors, temp/humidity, etc rarely are) - so those will always risk being out of sync after a power outage.

4 Likes

The question/point was:

Of course, this could be done in SW, no doubts.
I gave an idea how to use a relatively simple and relatively cheap solution
how to create an power outage event say, for the RM or whatever Custom App
or even platform event. This assumed the HE is powered from the UPS.

Yes a "Hubitat Resilience Kit"! Consists of a $100 UPS, A $50 pre-made contact/relay for power status detection. A built in app to detect power outages and respond appropriately for a low subscription price of $20/month and a disclaimer that they don't guarantee that it will handle every/any scenario that you automate.... So what was a $120 hub now costs $270 plus $240 a year to operate.... I wonder what kind of B!T@H!N6 that would cause in the community LOL. No thanks. I enjoy the challenge of making my own to meet MY needs.... Most of the tools are there....

2 Likes