Fire alerts
If a sensor determines with a high probability that a smoldering fire has occurred then fire alerts are immediately sent to notify users of the fire.
When the Silvanet system detects a fire, fire alerts are immediately sent to registered users via email and the Site Management app.
Warning - Let Sensors calibrate for 14 days, otherwise false alarms occur
Allow sensors to calibrate for 14 days after deployment. Silvanet sensors detect fires ONLY after the sensors have been calibrated (adapted) to the environment in which they have been deployed. Sensor calibration requires 14 days.
Consequently, before the calibration period ends, the sensors do not provide accurate readings and may generate false alerts.
How sensors generate fire alerts
If the sensor detects a decline in air quality, it triggers a fire detection process to determine if the observed deterioration is from a smoldering fire or from some other source.
If it determines with a high probability that the deterioration is from a smoldering fire, then the sensor sends fire alerts to the Silvanet Mesh Network which is forwarded to the Silvanet Cloud for distribution via email and the Site Management app.
A smoldering fire is defined to be a slow, flameless combustion of a biomass material such as forest floor material, branches or leaves.
After a sensor detects a decline in air quality, the sensor needs only 2-3 minutes to determine if a fire is present:
The gas sensor runs a set of gas scans.
Each gas scan is compared to a Fire Model.
If the gas scans determine a high probability of a fire, then fire alerts are triggered.
Fire alerts are sent to the Silvanet Cloud.
Silvanet Cloud sends fire notifications to registered users.
Returning to normal values after tests
After the sensor performs gas scans, it needs approximately 15 minutes for the sensor to return to baseline. During this time, the sensor does not perform fire detection.
1. Determine normal air quality
After the Wildfire sensor has been calibrated, the gas sensor continuously monitors the microclimate of the forest air to measure air pressure, temperature and humidity. The sensor also "smells" the air around the sensor for the presence of Volatile Organic Compounds (VOCs) and carbon monoxide.
VOCs are compounds that have a high vapor pressure and low water solubility.
Continuous monitoring results in the sensor learning to distinguish between Normal Air Quality and Declining Air Quality:
Normal Air Quality: The result of the 14 day calibration period of the sensor. It is the baseline reading of the air quality. This process continues after calibration.
Declining Air Quality: Could be the result of fumes from a smoldering fire but it could also be the result of gasses which the sensor has detected such as diesel fumes from a passing truck, cigarette smoke or other factors that can cause a decline in air quality.
The sensor “smells” the air on a regular basis to determine if the air quality has deviated from the saved Normal Air Quality. Air quality is quantified as an Index of Air Quality (IAQ) which has a range of values that indicate the quality of air within range of the sensor.
2. Decline in air quality occurs
Once per minute, the sensor checks the air quality of the microclimate near the sensor. If the air quality has significantly deteriorated within this short period of time (using a default sliding time window of 5 minutes), then this is considered a deviation from normal air quality.
When the sensor detects a a decline in air quality, this decline triggers a prescribed set of gas scans in the sensor's built in Bosch gas sensor. During each gas scan, hydrogen, carbon monoxide and other Volatile Organic Compounds (VOCs) are measured.
3. Gas scans compared to a ML Model
For each gas scan, the sensor compares the results of the gas scan with with pre-trained Machine Learning (ML) Models. This is done to determine if the observed decline in air quality can be classified as wildfire smoke or from some other source. As these gas scans consume a lot of energy, they are kept to a minimum.
These models have been trained in Dryad's lab to detect typical gas compositions of smoke from burning trees.
ML (Machine Learning) Models are pre-trained in the Dryad laboratory and can be fine-tuned / programmed specifically for the species of trees present in a particular deployment. Updated models can be remotely installed in the sensors by Dryad’s support team using FUOTA.
Firmware update Over the Air (FUOTA) is a standard for distributing firmware updates using unicast or multicast. It allows firmware updates to be delivered to many devices (Silvanet Sensors) at the same time efficiently and securely.
Low probability of fire: If the sensor determines some other source has caused the decline in air quality (diesel fumes or some other similar gasses), then the sensor stops the gas scans and does not trigger a fire alert. In this case the sensor must be reset which takes approximately 15 minutes.
High probability of fire: If the sensor determines that the decline in air quality has a very high probability of being the result of a smoldering fire, then gas scans are terminated and the sensor immediately transmits a fire alert to the Silvanet Mesh Network.
4. Fire alerts sent to Silvanet Cloud
A fire alert initiates a packet stream in which fire alert data packets are sent via the Mesh Network to the Silvanet Cloud. These packets include:
Device ID of the sensor
Timestamp
Location of the sensor
Other details
Fire alerts sent via email
When a sensor detects a fire, a fire alert email is immediately sent to registered users who can then act to extinguish the fire.
Total alerts: Total number of fire alerts sent from the detecting sensor.
New alerts in 2 hours: Number of alerts sent from the detecting sensor within the last two hours.
Last sending alert sensor: Includes the date and time of the fire alert, the Sensor ID of the detecting sensor, GPS location (Latitude and Longitude) of the sensor and a link to a map view of the fire location.
By selecting the Map link, Google Maps opens to show the location of the sensor that detected the fire. If Google Maps is not installed, it launches a browser on the smartphone and opens Google Maps in the browser.
Fire alerts in Site Management app
When a sensor detects a fire, the same fire alert that is sent as an email also appears as a fire alert in the Site Management app.
Selecting Alert Center, all fire alerts are displayed. Expanding the fire alert notification shows details of the detected fire.
Total alerts: Total number of fire alerts sent from the detecting sensor.
New alerts in 2 hours: Number of alerts sent from the detecting sensor within the last two hours.
Last sending alert sensor: Includes the date and time of the fire alert, the Sensor ID of the detecting sensor, GPS location (Latitude and Longitude) of the sensor and a link to a map view of the fire location.
Location of sensor: Selecting the Arrow icon next to Location opens Google Maps which displays a pin at the location of the sensor that detected a fire.
Mute selected fire alerts
Fire alert notifications from specific sensors can be muted.
From the Alert Center, select a Site generating fire alerts.
From the Site, select a sensor generating fire alerts.
Select the checkbox next to the Sensor name. The Mute notification icon appears.
From the dropdown choose the reason for clearing the alert:
Test Alert: Alert sent during testing the deployment with a controlled burn.
False Alert: Alert sent during the sensor calibration phase.
Fire Extinguished: Detected fire that caused the fire alert to be sent has been extinguished.
MQTT interface for fire alerts
An MQTT interface is available for 3rd party alarm interfaces / apps.
For more information, contact Dryad Support.
Last updated
