Sensor calibration

A deployed Silvanet sensor requires 14 days for the low-power, built-in BME866 gas sensor to calibrate, that is, to "settle in". This ensures the gas sensor adapts to the target environment. After 14 days, the gas sensor environmental readings eventually stabilize to an optimal level of sensitivity.

Calibration process

When a sensor is delivered to the user and deployed, it is ready to be charged and calibrated. No additional settings need to be made to the device.

• 1 day to charge the sensor after deployment

  After the sensor is installed on a tree and synced to the Silvanet Cloud (using the Dryad Deployment app), 1 day is required to fully charge the supercapacitors. Once charged, the sensor’s solar panel continuously charges the supercapacitors, except during the night.

• 14 days to run the calibration

  After the sensor has been fully charged, it begins a calibration process which typically takes 14 days. A Machine Learning (ML) algorithm running in the sensor performs this calibration. The sensor runs a gas scan for several days and uses the results of this gas scan to build an initial model of the environment air quality. It then deploys this model and sets an auto-trigger configuration. This ensures any false-positives are eliminated (sending fire alerts when there is no fire) as well as optimizing the sensors for fire detection.

• Sensor ready for fire detection

  After calibration, the sensor is ready to detect fires.

• Calibration process is repeated if sensor is moved

  If the sensor is moved to a new location, it automatically runs the calibration process again.

Continuous monitoring

After calibration, the gas sensor continuously monitors the microclimate of the forest air to measure air pressure, temperature and humidity.

It also "smells" the air around the sensor for the presence of Volatile Organic Compounds (VOCs) and carbon monoxide. By continuously monitoring the microclimate, gas sensor tracks historic readings of VOCs to produce a stable baseline called Normal Air which is unrelated to the presence of VOCs associated with a smoldering fire. This produces an Index of Air Quality (IAQ) value.

Normal and declining air quality

Continuous monitoring results in the sensor learning to distinguish between normal air quality and declining air quality. A decline in air quality is reflected as an increase in the IAQ value.

This increase in the IAQ value, which is a deviation from normal air quality, triggers a process to determine if the deviation is the result of gasses produced by the smoldering phase of a burning tree. This begins with a Phase 1 detection process.