Display "resets" every so often, loses loads, always on breakout

So, I’ve got a “spare” iPad that I’m using as a “monitor” here on my desk, which is running the iOS Sense app. Most of the time, the display is pretty sensible.

Every so often, the “bubble” part of the display blanks, out, turns to the “chasing dots” display, and redraws. When it does this, it almost always loses identified loads (eg AC) and switches those to “Other”…and all my smart plugs, which were listed under Always On, get separated out.

It’s almost like the Sense is crashing and rebooting (or there’s a watchdog that restarts a process).

Any idea what might be going on, or how I can fix it?

How many smartplugs do you have ? And how solid speed-wise is your internet, LAN, WiFi ? I don’t think it’s a crash, but more likely the Sense monitor having a challenge communicating with all the smartplugs, then back to the Sense mothership, then updating the app

It does something like this:

  • Sends a broadcast every 2 seconds to all smartplugs to report in
  • Smartplugs all respond unless some of the packets collide - not sure there’s even retransmission for collisions in the smartplug protocol.
  • Sense monitor collects all responses and sends back to the Sense mothership - I’m sure the monitor has some kind of deadline for receiving responses.
  • If the monitor misses a deadline or comes back with incomplete data, then the Sense updates the app with missing smartplugs.

Combine lots of smartplugs with network traffic and there are probably going to be some misses. I would contact support to see if there are ways to optimize.

I have four Kasa strips and something like three plugs. It’s not a lot.

My network is extremely fast, gigabit to the pole, all reliable/fast internally. I’m a software developer who works at home, and I’m super aware of networking issues. I’m pretty certain that this isn’t the network.

It just happened again while I was typing this - the bubbles all disappeared, the “wait” animation came on, and then the smart plugs came back with a single giant other (3800+W). It’s not that the plugs are missing: in fact, the plugs are reliable. It’s the rest that’s not.

Another thing that was happening: my espresso machine is on a strip. It’s obviously a large load when brewing. When it turns off, it drops to <6W. But Sense still showed it as a load, of 1225W, even though Kasa showed it as 5W. And it came up as “Breadmaker” in the list, even though the load was reported via the plug as a specifically measured load.

It’s weird behavior…

You are probably close, if not above the limit where the number of smartplugs can affect response times. I was there too with 3 x HS300s plus 12 HS110s. Contact support and they may be able to give some guidance.

From the FAQ
Is there a limit to the number of smart plugs I can connect to Sense?
The number of smart plugs is limited by Wi-Fi network bandwidth. Typical Sense users can use at least 20 smart plugs without issue.

Yeah, I saw that, and also the indication that an HS300 was “more than one but less than six”. So I don’t think that’s the issue here. I don’t see any problems at all with the actual plug-related loads. They’re responsive and present. Plus, this happened when I had fewer strips (I started with two), and it hasn’t gotten worse.

Also, interestingly, leaving the app open definitely causes more transfer from the Sense to the network (although I don’t know if it’s going outside my local network). In the last 24 hours, the monitor has done nearly 1GB of data I/O, which is more than I’ve seen before…

A while back some of us were noting that data dropouts occurred more frequently when the app or web app was invoked and also when open. There might be some live updates straight-ish from the monitor to the the bubble or to the power meter when open, that cause extra upstream traffic.

But I wouldn’t write off smartplug traffic as an issue - take a close look at your power meter and smartplug device power meters for dropouts, places where the data value drops to zero fro short periods in the middle of an active waveform.

I’ve seen occasional dropouts since well before I brought in any smart plugs (I’ve had the sense for years). In fact, this morning, there was one between 6:56am and 7:14am. But the network was up and there’s no real explanation for it. There are other occasional “burps”, too…but, again, that’s not anything new afaict.

I would definitely suggest writing into Support here. This does, from the surface, sound like it’s connection related.

https://sense.com/contact

I’ll write in to support - thought you guys might have more ideas here.

Note that I’ve got an access point (“pod”) right next (quite literally) to the Sense to ensure it has good connectivity. I’d wire it in if I could, given the relatively ancient WiFi chipset/stack, but that’s not an option. And the device logging from the WiFi controller shows no connection failures or reconnects, which is why I’m pretty sure it’s not connection related.

Also - why would a connection failure/burp cause what I’ve described? That is, a load that gets “un-identified” and all of the items that used to be “part of” Always On to get removed from there, and seemingly all the data about their idle/active states removed? That must be resilient to network burps, should there be one, no?

(Not trying to be argumentative, so I hope this isn’t coming off that way – just trying to understand the behavior.)