This program parses Ruuvi packets (RAWv2) from standard input and prints their payload as JSON to standard output.

Get a ready release or use the Go toolchain like this:

$ go install github.com/susji/ruuviparse@latest

We can feed Ruuvi's test vectors to ruuviparse:

$ echo '0512FC5394C37C0004FFFC040CAC364200CDCBB8334C884F' | ruuviparse | jq

{
  "Type": 5,
  "Timestamp": "2024-04-04T00:20:22.201683+03:00",
  "Temperature": {
    "Valid": true,
    "Value": 24.3
  },
  "Humidity": {
    "Valid": true,
    "Value": 53.489998
  },
  "Pressure": {
    "Valid": true,
    "Value": 100044
  },
  "AccelerationX": {
    "Valid": true,
    "Value": 4
  },
  "AccelerationY": {
    "Valid": true,
    "Value": -4
  },
  "AccelerationZ": {
    "Valid": true,
    "Value": 1036
  },
  "BatteryVoltage": {
    "Valid": true,
    "Value": 2.977
  },
  "TransmitPower": {
    "Valid": true,
    "Value": 4
  },
  "MovementCounter": {
    "Valid": true,
    "Value": 66
  },
  "SequenceNumber": {
    "Valid": true,
    "Value": 205
  },
  "MAC": "cb:b8:33:4c:88:4f"
}

Let us assume that you have a fleet of Ruuvi sensors and you would like to dump their (valid) temperature readings as their data is somehow streamed into a MQTT broker. Further, let us assume that

• there is a MQTT broker running at mqtt.example.com
• some kind of Ruuvi sensor data is being published in topic ruuvi/001
• you have a CA certificate, client certificate, and client private key
• you can dig out the BLE Advertising Payloads with jq from an array called ads...
  • which contains zero or more objects each containing hex-encoded Advertising Data in ad like this, and
  • the manufacturer-specific data is as described in the Ruuvi documentation.
• you have a suitable MQTT client installed
  • the example below uses mosquitto_sub

Your data shape would be something like this:

{
  "ads": [
    {
      "mac": "001122334455",
      "ad": "0201061bff990405..."
    },
    {
      "mac": "AABBCCDDEEFF",
      "ad": "0201061bff990405..."
    }
  ]
}

The Ruuvi RAWv2 Payload begins at 05 above, so from 7th octet which is index 14 in the hex-encoded string. We can also pick Ruuvi messages by filtering with ff9904 which means manufacturer-specific data and Ruuvi's unique identifier. To parse and pretty-print the temperature values, you can do something like this:

$ mosquitto_sub \
      -h mqtt.example.com \
      -t ruuvi/001 \
      --cert ruuvi-listener.cert.pem \
      --key ruuvi-listener.private.key \
      --cafile ca.crt \
      -i ruuvi-listener \
      | jq --raw-output --unbuffered \
            '.ads[].ad | ascii_downcase | select(.[8:14] == "ff9904") | .[14:]' \
      | ruuviparse \
      | jq --raw-output --compact-output \
            'select(.Temperature.Valid) | [.Timestamp, .MAC, .Temperature.Value] | @tsv'

For an example how to insert the data into a database, see the documentation for ruuviscan.

For a minimalistic example on how to expose Ruuvi values in Prometheus' text-based exposition format see scripts/ruuvimetrics.sh. The idea of the script is this: We will again use mosquitto_sub to obtain Ruuvi messages, parse and output them as JSON with ruuviparse and then generate a Prometheus-compatible metrics file of the sensor values. In addition to this script, you will probably need a way to serve the generated file over HTTP so Prometheus or a similar tool can periodically read its contents. After this, it's trivial to consume the sensor values with something like Grafana.

In the script, you will want to modify the broker settings and the ACT variable to copy the generated metrics file to your intended TARGETDIR. The resulting file might look like something like this:

# Generated at Sun Apr  7 14:36:07 UTC 2024 by ruuvimetrics.sh
# TYPE ruuvi_temperature gauge
ruuvi_temperature{mac="aa:aa:aa:aa:aa:aa"} 5.185 1712500560000
ruuvi_temperature{mac="bb:bb:bb:bb:bb:bb"} 2.615 1712500546000
ruuvi_temperature{mac="cc:cc:cc:cc:cc:cc"} 2.865 1712500567000
ruuvi_temperature{mac="dd:dd:dd:dd:dd:dd"} 22.289999 1712500567000
# TYPE ruuvi_voltage gauge
ruuvi_voltage{mac="aa:aa:aa:aa:aa:aa"} 2.929 1712500560000
ruuvi_voltage{mac="bb:bb:bb:bb:bb:bb"} 2.865 1712500546000
ruuvi_voltage{mac="cc:cc:cc:cc:cc:cc"} 2.943 1712500567000
ruuvi_voltage{mac="dd:dd:dd:dd:dd:dd"} 3.038 1712500567000
# TYPE ruuvi_movement gauge
ruuvi_movement{mac="aa:aa:aa:aa:aa:aa"} 183 1712500560000
ruuvi_movement{mac="bb:bb:bb:bb:bb:bb"} 125 1712500546000
ruuvi_movement{mac="cc:cc:cc:cc:cc:cc"} 88 1712500567000
ruuvi_movement{mac="dd:dd:dd:dd:dd:dd"} 154 1712500567000

For testing the idea, something like this will suffice to serve the resulting metrics file over HTTP:

$ cd $TARGETDIR && python3 -m http.server 9200 --bind 127.0.0.1

If you think you need elevated privileges to run the script, you should instead modify your target directory privileges accordingly. For some error-tolerance, wrap the invocation in a loop like

while true; do
    echo "begin $(date)"
    ./ruuvimetrics.sh
    echo exited
    sleep 10
done

or use some kind of a process manager.