Fire Detection using IoT Devices
The IoT Device
The IoT Edge Device (IoTED) has been developed, within the context of SILVANUS, as a means for an early fire/smoke detection mechanism, as well as a monitoring device for rural areas. Unlike other (current) approaches, this device was designed to be power independent and most importantly to process the collected data on the spot (edge), to reduce any latencies and identify possible critical events as soon as possible.
In detail, the IoTED consists of a waterproof case, with a strap for fastening it on trees/poles, and encloses widely used sensors and components. It is powered by a power bank that will be recharged with the help of solar panels, making it fully power independent as electricity is not always available. Moreover, the device is responsible for the collection of sensory data (e.g., humidity/temperature), capturing images of its surroundings, but most importantly their processing on the spot. Pre-trained fire and smoke detection Machine Learning (ML) models are employed (on the IoTED) for the assessment of possible fire/smoke events in the newly captured image. In the case of an identified fire/smoke event, the device sends a warning message to the SILVANUS framework, along with complimentary data (i.e., geo-location), to inform the first responders (e.g., firefighters) to take any necessary actions.
Potential benefits of the use of the IoTED, among others, are:
- Dynamic monitoring of remote areas.
- Constant checks for fire/smoke events within the area of interest, using ML.
- Close monitoring of actual fire/smoke events.
- Identification of fire-prone areas.
- Historical database/documentation of areas.
- Development of other monitoring/preventive models (e.g., fire spread model), using the collected data.
- Assessment of a fire’s severity.
- Improvement of use and allocation of a station’s resources (e.g., fire tracks).