davidp11
@david-patel · Joined March 2025
Facility manager · hospitality · new to IoT
- Topics
- 8
- Replies
- 62
- Points
- 0
Forum Replies Created
We went through the same thing at our hotels last year, so here is what I learned (I’m not an engineer, so this is the manager version).
Edge is software on the licensed gateway. It reads your Modbus devices on site, and each meter or controller shows up in Hermetry as its own child device. The gateway dials out over TLS, nothing on your site has to accept incoming connections.
If the broadband drops, licensed gateways can keep the readings locally and send them when the link is back. Each reading carries the time it was taken, so the history fills in at the right place. How much it can hold depends on the model, so ask for the figure for your model before you rely on it. Our integrator had to check that for us.
What surprised me was that alarm rules and flows run in Hermetry, on the cloud side, and the box only reads and buffers. So during an outage nobody gets an alarm e-mail, and when the backlog arrives the alarm rules look at the late readings by their own rules. Anything that must stop a press in a fault stays in the PLC or the controller, never in the cloud. Our maintenance lead was happy once he heard that.
The page that settled it for us: Licensed gateways and Hermetry Edge
Our maintenance guy did the ladder round last month with the OTA-capable firmware. Yesterday we pushed 2.1.1 (the reconnect fix) to my office board first, then to the profile in waves of 5. All 30 show Updated, no ladder. One room went “No answer in time” because the board was unplugged by housekeeping, it updated the next morning when it came back. The first upload was refused because the Name was “Room-Sensor” with capitals and the boards report “room-sensor”.
I’m not an engineer, but this happened to us last year, so here’s what I learned the expensive way.
We bought 12 leak sensors for our hotels from a small vendor, LTE-M, certified module, all reporting to Hermetry. Three months in, all 12 went offline on the same morning. I only noticed because the Device offline alarms fired for every one of them within ten minutes, before breakfast. The vendor spent a week on it. It turned out the operator had started rejecting devices whose IMEI range (the first 8 digits, the TAC) wasn’t registered with them as an approved device. The module had its approvals, but nobody had ever registered the finished sensor.
What fixed it: the vendor went through the operator’s device approval, which for a product built on a certified module was mostly paperwork plus a short field test. Our TAC got added and the sensors came back the next day with no changes on our side.
So for your 300 trackers, I’d call the operator now, before the build, ask what their approval looks like for a product on a certified module, and get your TAC registered. And set up a Device offline rule on the tracker profile. Your trackers wake every 30 minutes over HTTPS, so I’d set the profile to Recent message with something like 1,800 s times 2 and a short offline delay on top, so you hear about a block the same morning.
Have you turned the little blue screw on the back? Mine showed blocks until I did. Not sure it’s the same thing.
Good to know, we have two of these old shields in the hotel laundry and IT is talking about new switches.
Thank you emily! You were right, I measured it: the bank cuts out about 30 seconds after the first uplink. Mine has no always-on mode, so I moved to 3 x AA on the 5 V pin and took the LED off. It has sent 140 uplinks since yesterday. Next step is a Pro Mini like you said.
Following, I have a similar setup planned for our hotel greenhouse.
We have something similar with a speaker in our hotel lobby, it buzzes when someone’s phone rings near it. Same thing?
115200 and the rn were it, I got OK on the first try after changing both. Joined on the second attempt (first one I’d typed the AppKey with a space in it on the module side). Bytes are showing in the frames log. Now reading up on codecs like Owen said.
Could it be an accumulator? We have those on our hotel’s booster pumps, they look like a round thing on a line.
Update after two weeks: my nephew helped with the Python part. The simple rule (CO2 rising + motion in last 30 min) was right 87% of the time against the front desk log, the tiny model 91%. It runs on the Uno, 24 weights. We’re not buying the vendor thing. And yes, I made a Read-only key like Brian said, my first one was an admin key…
I have the flapping alarm problem with a UPS battery at our hotel, following this.
Didn’t know about Allowed networks, adding it to our hotel profile too.
Wouldn’t Tor slow everything down a lot? We tried a VPN in our hotel and the door sensors got laggy.
Following, I want to do something similar for meeting rooms in our hotel.

