<aside> 💡 THE STARTING POINT:

detected Anomaly example flow.json

</aside>

1. You need to decide which “error”-messages you want to store as detected anomalies. Use their MQTT-IN nodes. E.g. for:

   • errors
   • cleaning
   • heating
   • emergency turn-off
   • material defects
   • etc.

2. After that place a json for the convertation from string to object:

   

3. In case the value from the machine/sensor comes in as a boolean value. You need to transfer it into a boolean number (1 or 0 instead of TRUE or FALSE):

   msg.payload.ERR_TEMP_FÜHLER = msg.payload.ERR_TEMP_FÜHLER ? 1 : 0;
   return msg;
   

4. After that. use a rising edge to only let values through that reach the 1 (rising edge = 0.999).

1. Finally, we add a TO JSON function to give the message a timestamp and give it a name according to the anomaly. After this, we only need to send the message by using the /detectedAnomaly MQTT-OUT node.

   //TO JSON function 
   msg.payload = {
   "timestamp_ms": Date.now(), //timestamp
   "detectedAnomaly": "ERR_TEMP_FÜHLER" //name
   }
   
   return msg;