DEVELOPMENT OF A NEW AUTONOMOUS ANTI-FIRE SURVEILLANCE SYSTEM, N + I (No more fires)
Early detection of fires is a fundamental need to preserve forests.
Although important advances have been made in this field, current systems have two major handicaps:
– The technology used is not the ideal one.
To date, they are based on image analysis with algorithms based on colorimetry analysis and / or on the analysis of information captured by thermal cameras.
– The systems cannot be installed anywhere in geography.
They must be installed at points where there is electricity supply, mobile or fixed data network coverage and where a foundation can be made to support a tower that supports the cameras.
Unfortunately, in our forests, both electrical and communication infrastructures are not accessible over large areas of land. Therefore, the current systems cannot be applied to these areas, since they are based on the sending of information to the control center, and they require network infrastructure and power supply.
Taking into account the above, the main objective of this project has been based on developing an early fire detection system with autonomous operation, both at the level of the electrical network and the communications network.
Our system is based on the analysis of images using Deep Learning technology in which, in addition to analyzing the colorimetry of the image, amorphous forms that can change shape are analyzed and the speed with which they change and move is analyzed, to stop This mode can detect fire and smoke and be able to differentiate smoke from clouds without false alarms.
There are systems that analyze shapes and their movement, but to our knowledge there are no systems that can analyze amorphous elements such as fire or smoke and their movement. This is a resounding success of our research work.
In addition to the new and unique image analysis method with which to achieve truly automated early detection, Aicox has developed the different subsystems to be able to install surveillance posts anywhere in the geography and give the system an autonomous character:
Thus, we have our own design of a 30 m high mast that can be installed without the need for a foundation, raising its sections of 1 m sequentially with a hydraulic lifting system.
We have photovoltaic power solutions, the sizing of which is parameterized by Aicox taking into account the irradiation in the area where the system will be installed.
Aicox has adapted image capture systems and satellite communication systems to optimize system performance and thus allow images and alarms to be sent anywhere in geography.
We have also designed an alarm monitoring and control system, which focuses on the management of system alarms. In it, the images of the cameras of the different local posts are collected (videowall) and their location is displayed on a map (Map). When an alarm is received, the images of the local posts that generate the alarm are displayed on a screen (video alarm) while in the Map application the alarmed posts are highlighted and the direction in which the fire has been detected is indicated .
If there are enough local poles, located correctly, the exact point of the fire could be determined (although this is not the subject of this project).
Indicate that we designed a mobile image acquisition and analysis unit to obtain training videos for our analysis system, which can also be commercialized to be used by our clients.
To commercially promote our solution, we have designed a datasheet ( download here ) and a commercial video that shows the installation of our system in a real scenario.
The project, named “ DEVELOPMENT OF A NEW AUTONOMOUS ANTI-FIRE SURVEILLANCE SYSTEM, N + I (No more fires) “ , and with file number IDI-20180924 , has been co-financed by both the CDTI (supported by the Ministry of Economy, Industry and Competitiveness), and the Smart Growth Operational Program – ERDF and has an approved budget of € 724,372.
