I need to get temperature readings for a school building (concrete, 4 floors) and trying to decide how to better design the network. Wifi or Zigbee.
Wifi:
I have purchased several ESP8266 with DHT-11 sensors, which are supper tinny, but they still need to be connected to power. And this being a school, don’t quite like the idea of the phone charger and power cord hanging out of each plug.
Zigbee:
I do like the Xiaomi sensors, but don’t know how to build a reliable Zigbee network in this type of building. Worried about it being too big of a space to cover for one hub. Or shall I add additional hubs?
I wanted to send this over to the forum for recommendations in this situation.
One alternative would be to use inexpensive Particle Mesh devices which are programmed in C++ … like the Arduino. Mesh might help if you have weak or no WiFi coverage in some locations that you want to monitor. It would also keep your nodes from consuming IP addresses on the WiFi network. Particle gives you multiple networking options. These include their mesh network, WiFi, and cellular. They also have an ethernet feather-wing for places where wired networking is available … and there is a POE option. The devices can also connect directly to a LiPo battery for cordless applications.
Code can easily be ported between Arduino and Particle, so the learning curve would be minimal. Particle also has a cloud which includes diagnostic tools and the ability to flash nodes that are already deployed. The DHT-11 sensors would also work with Particle devices.
I use Particle devices with MQTT to interface with HA, so day-to-day operation is local, but I log data to Google Sheets, and use some custom HTML to interface with some Particle devices via the Particle Cloud. I also have one Particle device webhooked to PushBullet so I get notifications on my cell phone.
Particle, the company, is focused on helping people develop IOT products for fun or profit. If you build a prototype or two, you can use their resources to have a custom board made that includes the MPU, the temperature sensor, and whatever else you might want. This could result in a much smaller sensor device in a 3D-printed case. Perhaps you can turn your project into something that you can market to other schools.
I was not aware of the particle mesh network. That is a very neat idea with self healing mesh network.
Looking through the devices, the particle devices seem to be similar to the devices using 5V power and ESP for network connections that we were considering. In the case of the school, we have multiple access points, so the network is not necessary the issue, but the safety of the endpoints themselves that need to be in the various classrooms.
I’ll read more on the particle website as it seems that the endpoints can be battery powered and put to sleep after transmitting, saving battery life, similar to the the Xiaomi sensors I was researching.
Have you considered using RF sensors operating at 433MHz?
The sensors are inexpensive and powered by batteries. They communicate directly with the receiver (i.e. not a mesh network) and over a fairly long distance. In good conditions line of sight can be 300 feet but I would assume considerably less for indoor use (100 ft?).
Probably the easiest way to receive data from these sensors is via an inexpensive Software Defined Radio (SDR) and rtl_433 software. It runs on a Raspberry Pi and the received sensor data can be published to MQTT for use by Home Assistant. Nooelec makes a quality SDR and is available with a range-boosting antenna for 433MHz (Amazon). There are off-brand knock-offs available on eBay etc but I’d go with a quality SDR to avoid surprises.
The sensors can be off-brand devices sold on AliExpress, etc or the garden-variety Acurite sensors sold by Amazon and elsewhere.
rtl_433 plus an SDR provides a 433MHz solution with the least amount of tinkering. However, if you like building/programming things then you can opt for other 433MHz receivers like openMQTTGateway running on an Arduino or ESP8266 device. Alternately, it can run on a Sonoff RF Bridge (i.e. replaces the RF Bridge’s stock firmware).
I was looking for a case for one of my projects, and I thought of your project when I saw this family of cases from Hammond. They could be screwed to the wall and shouldn’t attract too much attention. It would also be easy to swap the plastic ends for wire mesh that would let air flow thru.
If you are considering Particle devices, I’d encourage you to post your project requirements there to see if others have done something similar … or how they might suggest meeting those requirements.
Also, feel free to message me here or there. I’d be glad to help with information and/or code. (I’m known as ‘bear’ in the Particle Community.)