Smart Meter Data Collector

The smartmeter-datacollector tool is a Python software which allows you to continuously retrieve data from supported smart meters (listed below). The acquired values can be forwarded to one or more data sinks like a MQTT broker.

This tool is an open source software funded by EKZ (Elektrizitätswerke des Kantons Zürich) and developed by Supercomputing Systems AG. The goal of this voluntary initiative is to create a tool that can read, decode and decrypt data (if necessary) from the local interface of various smart meters and convert it into human-readable and simple formats. This should eliminate the complexity of decoding DLMS and similar protocols for the user. The user has the data in a format, which he can easily use in various use-cases ranging from simple visualization (see Method 1: Raspberry Pi image with demo) to integration into energy management algorithms or into custom home automation software.

The following smart meters are supported (see Wiki/Home for detailed information):

• Landis+Gyr E450:
  Data pushed by smart meter over CII interface (wired M-Bus, HDLC, DLMS/COSEM)
• Landis+Gyr E570:
  Data pushed by smart meter over CII interface (wired M-Bus, HDLC, DLMS/COSEM)
• Landis+Gyr E360:
  Data pushed by smart meter over P1 interface (P1, HDLC, DLMS/COSEM only, no DSMR)
• Iskraemeco AM550:
  Data pushed by smart meter over P1 interface (P1, HDLC, DLMS/COSEM only, no DSMR)
• Kamstrup OMNIPOWER with HAN-NVE:
  Data pushed by smart meter over inserted HAN-NVE module (wired M-Bus, HDLC, DLMS/COSEM)
• Siemens TD-351x:
  Data fetched over bidirectional IR interface (IEC 62056-21, Mode C, unencrypted)

Note: All smart meters integrated so far push binary data encoded with HDLC (IEC 62056-46) and DLMS/COSEM. Both unencrypted and encrypted DLMS messages are accepted by the software. The P1-DSMR ASCII-based protocol is not supported.

The following data sinks are implemented:

• MQTT (v3.1.1):
  • Encryption
    • Unencrypted connection
    • Encrypted connection
      • TLS
      • optional custom CA certificate
  • Authentication
    • Unauthenticated
    • Authenticated with username / password
    • Authenticated with client certificate
• MQTT RL-DSP:
  like MQTT sink but with topic- and payload-format specified in VSE RL-DSP CH2024 document
• Logger to stdout

smartmeter-datacollector is fully configurable through a .ini configuration file. The smartmeter-datacollector-configurator web interface can help to create and modify the configuration.

• Smart Meter Data Collector
• How to install
  • Method 1: Raspberry Pi image with demo
  • Method 2: Python package
    • Python Requirements
    • Installation
  • Method 3: Debian package
    • Debian Requirements
    • Installation
• How to use
  • Configuration
    • Specification
    • smartmeter-datacollector-configurator
  • Run manually
    • Configuration
  • Run as a systemd service
    • Configuration
• How to develop
  • Development Requirements
  • Installation
    • Debian / Ubuntu
  • Checkout the code
  • Setup Development Environment
  • Custom Commands / Workflows
• Acknowledgements

How to install

There are different methods how to use smartmeter-datacollector.

1. Raspberry Pi image with demo
2. Python package
3. Debian package

Method 1: Raspberry Pi image with demo

For a very easy first time usage of smartmeter-datacollector we provide a Raspberry Pi image (based on Raspberry Pi OS) which contains the following parts:

• smartmeter-datacollector as a systemd service
• smartmeter-datacollector-configurator web interface
• Demo setup with:
  • mosquitto as a local MQTT broker to publish the measurements from smartmeter-datacollector
  • Telegraf to collect measurements published by smartmeter-datacollector over MQTT and store them in InfluxDB
  • InfluxDB to store the measurements
  • Grafana to visualize the measurements

See Wiki/Demo for detailed setup instructions.

Method 2: Python package

smartmeter-datacollector can be installed as a Python3 package either from PyPi or manually using a released wheel.

The Python3 package does not contain any infrastructure to run smartmeter-datacollector in the background or to automatically start it during a boot sequence. If such infrastructure is required either see Method 3: Debian package or provide it yourself.

Python Requirements

• Python >= 3.10 (tested with 3.13)

Installation

Install the package in a virtualenv manually or via using pipx

pipx install smartmeter-datacollector

Similarly the smartmeter-datacollector-configurator web interface can be installed with

pipx install smartmeter-datacollector-configurator

Method 3: Debian package

smartmeter-datacollector is also available as a Debian (.deb) package from releases which installs the application bundled with its Python dependencies in a zipapp / .pyz file. The Debian package includes a systemd service file which enables smartmeter-datacollector to automatically start after booting the system.

Requirements

• Distribution: Debian based (like Debian, Ubuntu, ..)
• Release: >= bookworm/12
• CPU architecture: independent

Installation

Download the Debian package from releases and install it with

sudo apt install ./python3-smartmeter-datacollector_*.deb

Similarly the smartmeter-datacollector-configurator web interface can be installed with a Debian package from its releases with

sudo apt install ./python3-smartmeter-datacollector-configurator_*.deb

How to use

The usage of smartmeter-datacollector depends on the installation method. Independent of the installation method a .ini configuration file is required to properly run smartmeter-datacollector.

Configuration

Manually create configuration

See Wiki/Configuration for more details on the available configuration options.

smartmeter-datacollector-configurator

To simplify the process of generating a valid .ini configuration for smartmeter-datacollector the companion smartmeter-datacollector-configurator web interface can be used. It supports

• a graphical approach to manage the configuration
• input validation to avoid invalid configurations
• loading / saving / discarding a configuration
• restarting smartmeter-datacollector (only if installed as a Debian package)

See Wiki/smartmeter-datacollector-configurator for more details on the web interface and the configuration possibilities.

Run manually

Run smartmeter-datacollector with the following command:

smartmeter-datacollector --config datacollector.ini

The following command line arguments are supported:

• -h, --help: Shows the help output of smartmeter-datacollector
• -c, --config CONFIG: Path to the .ini configuration file
• -s,--saveconfig: Create a default .ini configuration file
• -d, --dev: Enable development mode

Configuration

The configuration file can be located anywhere and use any filename. If no .ini configuration file is specified a default configuration with the following options is used:

• Landys+Gyr E450 smart meter in unencrypted mode connected to /dev/ttyUSB0
• LOGGER sink
• MQTT sink connected to a local broker without encryption or authentication

Run as a systemd service

When smartmeter-datacollector has been installed as a Debian package it provides a systemd .service file named smartmeter-datacollector.service. Therefore the service can be managed using the systemctl command. See Run as a systemd service for possible commands.

How to develop

Help from the community for the smartmeter-datacollector project is always welcome. Please follow the next few chapters to setup a working development environment.

Development Requirements

• Python >= 3.10, <= 3.14
• poetry
• Optional software packages
  • debhelper
  • dh-make
  • rename

Installation

Debian / Ubuntu

To install the listed minimal requirements run the following command:

sudo apt install git python3

Also install poetry according to Poetry - Installation.

To install the optional requirements also run the following command:

sudo apt install debhelper dh-make rename

Checkout the code

Use git to clone / checkout smartmeter-datacollector from GitHub using

git clone https://github.com/scs/smartmeter-datacollector.git

Setup Development Environment

smartmeter-datacollector uses poetry to manage its dependencies and setup a virtual environment. Run the following command to setup the initial development environment:

poetry install

This will install all runtime and development dependencies for smartmeter-datacollector in a new virtual environment. Now you are ready to start working on smartmeter-datacollector.

Custom Commands / Workflows

smartmeter-datacollector offers a few custom poetry run poe commands to simplify certain development workflows:

• build_py builds a Python package which can be uploaded to PyPI
• build_shiv builds a self-contained zipapp (.pyz) including dependencies (but without interpreter) using shiv
• build_deb builds a Debian package for the current development platform
• build_srcdeb builds a Debian source package which can be used to build a Debian (binary) package for any platform (e.g. using pbuilder)
• clean removes build output from the working directory
• debianize creates a debian/ directory used to build Debian source / binary packages
• format automatically adjusts the code to follow the autopep8 code formatting rules
• format_check checks if the code follows the autopep8 code formatting rules
• isort automatically re-orders the import statements using isort
• isort_check checks if the order of the import statements is correct using isort
• license makes sure every Python (*.py) file contains the proper license header
• lint automatically adjust the code to follow the pylint rules defined in pyproject.toml
• lint_check checks if the code follows the pylint rules defined in pyproject.toml
• test runs all unit tests using pytest

Make sure to run format_check / format, isort_check / isort, lint_check / lint, license before committing changes to the repository to avoid unnecessary development cycles. smartmeter-datacollector uses GitHub Actions to check if these rules apply.

Acknowledgements

smartmeter-datacollector and its companion project smartmeter-datacollector-configurator have been developed by Supercomputing Systems AG on behalf of and funded by EKZ.