Sector of activity
Without mentioning human losses, the cost of a forest fire essentially depends on three components:
The ADELIE automatic forest fire detection system detects the presence of smoke in the natural environment thanks to a unique principle of image capture and processing. It allows, from one or more high points and up to the horizon, permanently and over 360°, the surveillance, detection and localization of fire starts.
Locally developed information is transmitted in real time to the operational centre. A visual and audible alert indicates to the operator the existence of a smoke detection and the images necessary to validate the alert are automatically displayed.
This system is based on 5 main themes:
The shooting equipment is spread out on towers, pylons or high points and covers the wooded surfaces visible from this point to the horizon and over 360°.
Each high point is equipped with two detection cameras covering 360°.
A remote-controlled video camera, known as a "doubt removal" camera, completes this system (monitoring of the fire before the arrival of the first unit).
The analysis of the images thus obtained is carried out by comparing two images, taken from the same orientation, to highlight any appearance of smoke. This thanks to a certain number of exclusive algorithms, developed by PARATRONIC, which make it possible to compare a very great number of factors not visible to the naked eye.
This analysis is carried out in 3 phases:
Comparison of images: Images are compared to highlight any changes, i.e. any movement or displacement of an object or any appearance of smoke.
Advanced analysis, based on the use of different algorithms: Not only are the differences highlighted are screened according to their shape, size, distance, etc... in order to eliminate any element other than smoke as much as possible, but other algorithms, of the "machine vector support" type or not using automatic classifiers, working on the basis of parameters extracted from one or more images, complete this analysis.
Following this analysis, are transmitted, by FTP via radio, ADSL or any other media, to the control computer system:
The images thus received (17,280 images per 24 hours and per equipped site) are stored for 24 hours. All images are therefore transmitted, whether or not they contain detection. If an image and its linked file show a detection, an automatic check is carried out so as to trigger an alarm only if this detection has not taken place in an exclusion zone (an area where permanent smoke is known, for example a factory chimney).
Together with the images, the system sends the operational centre the angular orientation (azimuth) and the estimated distance of the detection from the high point.
All information (fire confirmation, contact details, etc.) is directly transmitted to the user's usual alert management system.
If only one tower has detected smoke, the telemetry distance is used. If at least two towers have detected the smoke, the control centre will confirm the location of the fire by triangulation.
Depending on the site, the transmission can also be carried out by any other high-speed communication media: ADSL, fibre, satellite, etc.
The existing towers or high points used are, depending on the topography and the transmission media used, connected in "mesh networks". The network is dimensioned to receive real-time camera views of all the watchtowers simultaneously.
This architecture makes it possible to cover large distances, in excess of 100 kilometres for radio solutions, by ensuring the necessary data rates for image transport. This transmission system can be secured in degraded operation by ADSL link.
At the operational centre, images of the various observation points and any alarms are received on one or more computer stations.
The "ADELIE" system allows for the efficient detection of any suspicious smoke and its precise location. In addition, it facilitates fire tracking and the guidance of fire-fighting means.
The system also makes it possible, without mobilising a unit, to quickly and accurately validate the location of a fire start signalled by a ¨APPEL 18¨
As with any automatic system, it is essential that there is human validation of the alarms transmitted. It is the operators of the operational centre who, with the help of a high definition camera equipped with a powerful zoom (26X), decide to validate the alarms and confirm the existence and importance of a fire outbreak.
Those responsible for monitoring operations can in turn remotely monitor the evolution of the situation thanks to these cameras, known as "doubt removal" cameras, without interrupting the detection system. The ADELIE detection system therefore remains fully active in order to prevent any eventuality of multiple fires.
The detected light(s) can then be viewed at the central station before the first unit is sent to the site. The number of units required can therefore be sent directly to the site.
During the intervention, ADELIE continues, through real time video, to give to the different persons in charge a great number of information on the evolution of the fire such as for example the extent and the colour of the plume of smoke, the view of the development of the event (which can be very variable according to the topology of the sector and the Weather) but also the live follow-up of the intervention of the canadairs and the result obtained .
Any system for monitoring and tracking natural hazards must operate under the worst imaginable conditions, including if these occur at the same time as a physical incident.
It is essential to ensure full continuity of service even in the event of equipment failure. All equipment must therefore be redundant and the switchover from failed equipment to replacement equipment must be automatic.
Each tower can therefore be equipped with one or two detection cameras, one or two computers and one or two transmission systems (e.g. Radio + ADSL).
A remote maintenance and remote control system is possible.
Implementation of the ADELIE system for detecting forest and natural area fires for the SDIS72
Functional cookies help to perform certain functions such as sharing website content on social media platforms, collecting comments and other third party functionality.
Performance cookies are used to understand and analyze key website performance metrics, which helps to provide a better user experience for visitors.
Analytical cookies are used to understand how visitors interact with the website. These cookies help to provide information on measures of the number of visitors, bounce rate, source of traffic, etc.
Advertising cookies are used to provide visitors with relevant advertisements and marketing campaigns. These cookies track visitors to the websites and collect information to provide personalized advertisements.
Other uncategorized cookies are those that are currently being analyzed and have not yet been categorized.
All cookies that may not be particularly necessary for the operation of the website and that are used specifically to collect personal data from users via analyses, advertisements or other embedded content are referred to as unnecessary cookies.