Using an inexpensive rtl-sdr dongle, it's possible to listen for signals from ERT compatible smart meters using rtlamr. This script runs as a daemon, launches rtl_tcp and rtlamr, and parses the output from rtlamr. If this matches your meter, it will push the data into MQTT for consumption by Home Assistant, OpenHAB, or custom scripts.
Tested under Raspbian GNU/Linux 9.3 (stretch)
Tested under Ubuntu 18.04.4 LTS
Install RTL-SDR package
sudo apt-get install rtl-sdr
Set permissions on rtl-sdr device
/etc/udev/rules.d/rtl-sdr.rules
SUBSYSTEMS=="usb", ATTRS{idVendor}=="0bda", ATTRS{idProduct}=="2838", MODE:="0666"
Prevent tv tuner drivers from using rtl-sdr device
/etc/modprobe.d/rtl-sdr.conf
blacklist dvb_usb_rtl28xxu
sudo apt-get install git
Install pip for python 3
sudo apt-get install python3-pip
Install paho-mqtt package for python3
sudo pip3 install paho-mqtt
Install psutil package for python3
sudo pip3 install psutil
Install Go programming language & set gopath
sudo apt-get install golang
https://github.com/golang/go/wiki/SettingGOPATH
If only running go to get rtlamr, just set environment temporarily with the following command
export GOPATH=$HOME/go
Install rtlamr https://github.com/bemasher/rtlamr
go get github.com/bemasher/rtlamr
To make things convenient, I'm copying rtlamr to /usr/local/bin
sudo cp ~/go/bin/rtlamr /usr/local/bin/rtlamr
Clone repo into opt
cd /opt
sudo git clone https://github.com/chrcraven/amr2mqtt.git
Copy template to settings.py
cd /opt/amr2mqtt
sudo cp settings_template.py settings.py
Edit file and replace with appropriate values for your configuration. Leave watched meters blank to discover all broadcasting meters in your area. If you want the service to wait until another process is running (ie Mosquitto MQTT) use the top command to find the process name and populate PROCESS_WAIT_FOR_START.
sudo nano /opt/amr2mqtt/settings.py
Copy amr2mqtt service configuration into systemd config
sudo cp /opt/amr2mqtt/amr2mqtt.systemd.service /etc/systemd/system/amr2mqtt.service
Refresh systemd configuration
sudo systemctl daemon-reload
Start amr2mqtt service
sudo service amr2mqtt start
Set amr2mqtt to run on startup
sudo systemctl enable amr2mqtt.service
Be patient! There is a built in delay of two minutes on service startup before data starts loading into MQTT. This delay is to ensure your machine is ready post boot. Use a GUI or command line tool to view the 'readings' topic on your MQTT server. I use a tool called MQTT Explorer which offers installation packages for Linux, Windows, and Apple. Download at https://mqtt-explorer.com/
You have made it this far and now have data feeding into your MQTT server. Now it is time to understand how this program works so you can make informed choices as to how you use this data. When amr2mqtt is run it first starts rtl_tcp, then rtlamr, and then enters into its main loop. rtlamr is set up to return SCM, SCM+, and IDM messages from broadcasting meters in a JSON format and the main loop of the program reads this JSON output from rtlamr and publishes it to MQTT using the following format:
readings/{meter_id}/{message_type}
For example, my electric meter broadcasts both SCM and IDM messages so the following topics would have JSON messages published to them:
readings/{my_meter}/scm
{
"Time":"2020-02-17T16:42:16.121571951Z",
"Offset":0,
"Length":0,
"Type":"SCM",
"Message": {
"ID":,
"Type":7,
"TamperPhy":2,
"TamperEnc":2,
"Consumption":974582,
"ChecksumVal":37576
}
}
readings/{my_meter}/idm
{
"Time":"2020-02-17T16:45:12.844042259Z",
"Offset":0,
"Length":0,
"Type":"IDM",
"Message": {
"Preamble":,
"PacketTypeID":28,
"PacketLength":92,
"HammingCode":198,
"ApplicationVersion":4,
"ERTType":7,
"ERTSerialNumber":,
"ConsumptionIntervalCount":235,
"ModuleProgrammingState":188,
"TamperCounters":"AgIAEgoA",
"AsynchronousCounters":0,
"PowerOutageFlags":"AAAAACAA",
"LastConsumptionCount":974588,
"DifferentialConsumptionIntervals:[6,6,5,9,10,7,5,4,8,9,9,9,5,5,6,11,13,8,5,4,6,9,8,8,7,6,4,5,4,5,5,4,3,3,7,8,8,3,5,5,4,9,12,7,4,3,3],
"TransmitTimeOffset":15,
"SerialNumberCRC":42424,
"PacketCRC":11799
}
}
For more information on the message types visit https://github.com/bemasher/rtlamr/wiki/Protocol. Message formats other than SCM, SCM+, and IDM are not currently supported by this fork.
This section assumes you have Home Assistant configured to subscribe to a MQTT server. This is my configuration from Home Assistant.
My electric meter records in watt hours and broadcasts both SCM and IDM message types. I keep the raw data from the meter in its own sensor and create a templated sensor which divides the watt hour sensor by 1000 to get the kWh reading. My utility bills by kWh.
#Electric Meter
- platform: mqtt
state_topic: "readings/{meter_id}/scm"
name: "Electric Meter Consumption"
unit_of_measurement: Wh
value_template: "{{ (value_json.Message.Consumption | int) }}"
- platform: template
sensors:
electric_meter_consumption_kwh:
friendly_name: Electric Meter Consumption (Billing)
unit_of_measurement: 'kWh'
value_template: "{{ (((states('sensor.electric_meter_consumption')|int)/1000) | float) | round(2) }}"
My gas meter reports in cubic feet and broadcasts the SCM message type. I keep the raw data from the meter in its own sensor and create templated sensors which report in CCF (100 cubic feet) and in therms. My utility bills by therm.
#Gas Meter
- platform: mqtt
state_topic: "readings/{meter_id}/scm"
name: "Gas Meter Consumption"
unit_of_measurement: ft3
value_template: "{{ (value_json.Message.Consumption | int) }}"
- platform: template
sensors:
gas_meter_consumption_ccf:
friendly_name: Gas Meter Consumption (CCF)
unit_of_measurement: 'CCF'
value_template: "{{ (((states('sensor.gas_meter_consumption')|int) / 100) | float | round(2)) }}"
gas_meter_consumption_therm:
friendly_name: Gas Meter Consumption (Therm)
unit_of_measurement: 'Therm'
value_template: "{{ (((states('sensor.gas_meter_consumption_ccf')|float)*('.98'|float)) | float | round(2))}}"
My water meter reports in cubic feet and broadcasts the SCM message type. I keep the raw data from the meter in its own sensor and create a templated sensor which reports in CCF (100 cubic feet). My utility bills by CCF.
#Water Meter
- platform: mqtt
state_topic: "readings/29677808/scm"
name: "Water Meter Consumption"
unit_of_measurement: ft3
value_template: "{{ (value_json.Message.Consumption | int) }}"
- platform: template
sensors:
water_meter_consumption_ccf:
friendly_name: Water Meter Consumption
unit_of_measurement: 'CCF'
value_template: "{{ (((states('sensor.water_meter_consumption')|int) / 100) | float | round(2)) }}"
As part of my installation I discovered that my meter readings would spike whenever I would restart Home Assistant. Since I like to tinker with my installation restarts are going to happen and to persist accurate readings across restarts I needed to use a custom component from https://github.com/tiaanv/hass-utility_meter_alternative to add a new option to the existing utility meter functionality. Follow the instructions from tiaanv. After installed follow the directions at https://developers.home-assistant.io/blog/2019/04/12/new-integration-structure.html to add a manifest.json and edit the domain in const.py to be 'utility_meter_custom'.
I have set up hourly, daily, weekly, and monthly meters for each 'billing' sensor.
#Utility meters
utility_meter_custom:
electric_meter_monthly:
source: sensor.electric_meter_consumption_kwh
mode: alt
cycle: monthly
electric_meter_weekly:
source: sensor.electric_meter_consumption_kwh
mode: alt
cycle: weekly
electric_meter_daily:
source: sensor.electric_meter_consumption_kwh
mode: alt
cycle: daily
electric_meter_hourly:
source: sensor.electric_meter_consumption_kwh
mode: alt
cycle: hourly
water_meter_monthly:
source: sensor.water_meter_consumption_ccf
mode: alt
cycle: monthly
water_meter_weekly:
source: sensor.water_meter_consumption_ccf
mode: alt
cycle: weekly
water_meter_daily:
source: sensor.water_meter_consumption_ccf
mode: alt
cycle: daily
water_meter_hourly:
source: sensor.water_meter_consumption_ccf
mode: alt
cycle: hourly
gas_meter_monthly:
source: sensor.gas_meter_consumption_therm
mode: alt
cycle: monthly
gas_meter_weekly:
source: sensor.gas_meter_consumption_therm
mode: alt
cycle: weekly
gas_meter_daily:
source: sensor.gas_meter_consumption_therm
mode: alt
cycle: daily
gas_meter_hourly:
source: sensor.gas_meter_consumption_therm
mode: alt
cycle: hourly