Fire detection and prediction equipment in forested areas.

Description

1 / 14 Fire detection and prediction equipment in forested areas.

[001] This utility model application refers to equipment for detecting and predicting fire in forested areas. The equipment consists of modules and individual codes, with the components connected to the board, code execution, and sensor interconnections, combining their codes to collect data with the objective of detecting smoke and fire in forested areas. The equipment is a proximity sensor, with an approximate detection capacity of around 3 m radius, enabling the issuance of alerts with the geolocation of smoke and fire detection so that contingency and incident response teams can mobilize and reach the site quickly and accurately to mitigate the impact, mainly to prevent the event from escalating, in terms of the environmental problem of fires. Fundamentals of the utility model

[002] Forest fires have become more frequent and intense, either due to population growth putting pressure on Conservation Units and occupying surrounding areas, or due to the expansion of new agricultural areas. One of the consequences of these factors has been the constant increase in large-scale fire outbreaks due to fires caused by intentional or accidental human actions. In this context, Conservation Units and Environmental Reserves have been significantly impacted. One of the main problems for effective firefighting is that many outbreaks take a long time to be detected, becoming visible later. Petition 870250056046, dated 02 / 07 / 2025, page 4 / 31 2 / 14 only in large proportions and with high destructive power, hindering any action by brigades or task forces in combating them. Added to this is the fact that several outbreaks occur in remote and difficult-to-access locations, making their detection difficult and therefore delaying arrival at the site for effective containment of the event when in its initial phase. This application aims for equipment composed of a hardware and software system, using the concepts of Internet of Things (IoT), Artificial Intelligence and remote sensing, that detects the presence of smoke and fire outbreaks, supported by additional environmental information, associated with an Artificial Intelligence (AI) algorithm.The equipment predicts the occurrence of fires and sends data on temperature, relative humidity, geolocation, and the presence of smoke and fire at the scene to firefighting units, enabling a prompt and efficient response, since the coordinates sent to the firefighting brigades allow them to reach the exact location of the incident. State of the Art

[003] The following state-of-the-art equipment and documents were found:

[004] Ineeds Forest Fire Detector (DIF): equipment designed to detect fire, carbon monoxide, and other gases from fire (not identified by the manufacturer). According to the manufacturer, its advantages include using artificial intelligence to eliminate false positives; IoT wireless communication for wireless data transmission; long-lasting battery; and the use of a solar panel for energy generation. Disadvantages include limited wireless communication for short periods. Petition 870250056046, dated 02 / 07 / 2025, page 5 / 31 3 / 14 distances, as it depends on interconnection equipment for long-distance connectivity, thus making its use unfeasible in areas more than 30 meters from the interconnection equipment; it does not store data locally; it does not offer geolocation of the fire location (not mentioned by the manufacturer); and it does not offer a smoke detection sensor.

[005] The MC-TENG equipment - developed by Michigan State University - United States aims to detect fire via temperature and carbon monoxide, sending warnings via alarm. Its advantages include: it can be recharged using air currents and sends data wirelessly to a database that retransmits to the control center via satellite. Its disadvantages include: high cost to establish the network and maintain the database; expensive satellite transmissions; it does not store data locally; it is still in the prototype phase and has not been field tested; and it does not offer a smoke detection sensor.

[006] Patent document CN1174755A refers to the technical field of forest fire prevention, specifically an intelligent radar monitoring early warning system. The subject matter of the present invention includes: a data collection unit, cloud database, radar density layout analysis unit, forest fire early warning unit and forest fire pre-analysis, as well as display alert units and terminals.

[007] The invention CN203415064U describes a utility model belonging to the field of fire alarms, in particular to an intelligent forest fire alarm positioning system based on a wireless sensor network. It consists of video monitoring equipment. Petition 870250056046, dated 02 / 07 / 2025, page 6 / 31 The intelligent 4 / 14 system, GPS positioning system, wireless sensor network, remote monitoring platform, fire pump, fire alarm display, and audible and visual fire alarm, with connected equipment, allows for real-time fire monitoring. In addition, the intelligent video monitoring device also includes an intelligent electronic temperature sensor, smoke detector, fire detector, and an intelligent electronic cigarette lighter.

[008] Patent CN216899001U discloses a utility model belonging to the technical field of environmental monitoring, in particular a device for monitoring the forest ecological environment. The monitoring assembly includes a monitoring box, a vertical plate, a temperature sensor, a humidity sensor, a smoke sensor, a PLC controller, and a signal transmitter. The monitoring box is located on top of the placement table, and the vertical plate is inside the monitoring box. The temperature sensor is located on the side wall of the vertical plate, the humidity sensor is located on the outer wall of the vertical plate, the smoke sensor is located on the side wall of the vertical plate, the PLC controller is located on top of the placement table, and the signal transmitter is located on the side wall of the vertical plate.

[009] The invention US2022398840A1 relates to a system and method for tracking forest fires over large areas. The invention concerns an improved firefighting method and system for smoke recognition, fire prediction, and fire growth monitoring. The system comprises a plurality of self-directed unmanned aerial devices (UAVs), comprising a UAV processor, memory Petition 870250056046, dated 02 / 07 / 2025, p. 7 / 31 5 / 14 of a UAV, a UAV communication device, a UAV database, and a server comprising a processor, secondary memory, a secondary communication device, and a secondary database, wherein the system is operable to: periodically obtain one or more first images of one or more regions of interest within a geographic area, using a plurality of self-directed unmanned aerial devices; allow obtaining information related to the ambient climate of regions of interest; obtain one or more land-related information items, wherein one or more land-related information items comprise characteristics of vegetation, land, topography, elevation, slope, and aspect for the geographic region; automatically map, using a deep neural network, one or more risk areas of one or more regions of interest, from one or more first images of one or more regions of interest;To periodically obtain one or more secondary images of one or more risk areas, mapped from one or more primary images of one or more regions of interest, using a plurality of self-directed unmanned aerial vehicles; to automatically recognize, using a deep neural network, one or more signals related to smoke or fire from one or more secondary images and one or more pieces of information related to ambient weather; and to automatically predict, through a deep neural network, the existence of a fire causing smoke, a fire, fire growth and spread based on one or more secondary images, one or more pieces of information related to ambient weather from one or more regions of interest and one or more pieces of information related to the Earth. Petition 870250056046, dated 02 / 07 / 2025, page 8 / 31 6 / 14

[010] Although the documents cited above reveal equipment with sensors for environmental monitoring, through devices for monitoring the forest ecological environment for fire prevention, they differ from the present utility model because it is an equipment that presents sustainability in the energization to power 16 gas, flame and fire detection sensors with connectivity via GSM and sending messages via SMS for 24 uninterrupted hours. Summary

[011] The differences between the equipment of the present utility model and the prior art are:

[012] Use of solar energy: it is designed to be self-sufficient in energy generation, using a solar panel that, during the day, sustains the energy supply of the equipment and at the same time supplies the battery for operation during the night and / or absence of sunlight (cloudy weather and / or rain);

[013] Use of four (04) carbon monoxide (CO) gas and smoke detection sensors connected to the Arduino microcontroller;

[014] Use of four (04) carbon dioxide (CO2) gas and smoke detection sensors connected to the Arduino microcontroller;

[015] Use of eight (08) flame and fire sensors connected to the Arduino microcontroller;

[016] Use of a (01) temperature, pressure and humidity sensor connected to the Arduino microcontroller;

[017] Data storage on SD flash card: data is stored / saved in case the connection becomes unavailable for later use; Petition 870250056046, dated 02 / 07 / 2025, page 9 / 31 7 / 14

[018] Connectivity with data transmission via GSM: transmission of temperature, pressure, humidity, smoke presence, fire presence and geolocation data in real time using SMS data transmission, which implies energy savings for the equipment compared to wireless transmission and the cost of additional equipment for wireless connection or mesh network equipment. Description of the Figures

[019] Fig. 1 - Microcontroller connection diagram with only the sensors and equipment components;

[020] Fig. 2 - Connection diagram of the microcontroller with sensors, modules and battery of the equipment;

[021] Fig. 3 - Appearance of the equipment with the various sensor modules;

[022] Fig. 4 - Hermetic box. Detailed description

[023] The testing phase on the Arduino was determined by the functional tests performed on the board, such as: connecting the cables to the board and the notebook; verifying the response via LEDs, power supply; uploading (transferring) the code to the board's flash memory; verifying the response to code providing a light signal output and verifying the response via serial output.

[024] The next phase consisted of assembling all the sensors, modules and codes individually, performing component connection tests with the board, code execution and, subsequently, performing the interconnections of all sensors, insertion of the respective libraries and modules, joining of their codes and subsequent data collection tests. Petition 870250056046, dated 02 / 07 / 2025, p. 10 / 31 8 / 14

[025] The description of the sensors, procedures and tests follows as shown below:

[026] 01 pin connected to the Arduino power supply; 01 pin connected to ground (neutral) at GND; 01 Serial Data (SDA) pin responsible for sending and receiving data and 01 Serial Clock (SCL) pin responsible for system time synchronization.

[027] The characteristics of the temperature and humidity sensor are: - Power supply: 1.5 V to 3.6 V; - Humidity measurement range: 0-100% RH; Temperature range: -40 to 105 °C; - Communication: I2C (SDA and SCL); - Humidity accuracy (10% RH to 95% RH): ± 2% RH Measurement time: 50 ms (milliseconds).

[028] After making the connections, the assembly was connected to a notebook via USB cable and the following steps were performed: initialization of the programming environment; insertion of the code; import of the temperature and humidity sensor library; error checking and debugging; sending the code to the Arduino; execution and verification of the data collected via serial output.

[029] The first sensor used in the initial tests was a temperature and humidity sensor, which performs temperature and humidity readings, with an operating voltage (potential difference) of 3.3 Volts (V) and low current consumption, between 2.5 milliAmperes (mA) during measurements and 100 to 150 microampere (μA) when in standby / rest mode. The humidity reading ranges are between 0 and 100% with a maximum error rate of 5%, and the temperature range is between -40 and 125 °C, with an error rate of + / - 0.5%. Petition 870250056046, dated 02 / 07 / 2025, p. 11 / 31 9 / 14

[030] The sensor was connected to the Arduino microcontroller and the operating code with the libraries was uploaded. Temperature and humidity readings were correctly taken with another device nearby to verify operation.

[031] Next, the smoke sensor was connected, which is responsible for detecting concentrations of flammable gases (LPG, Methane, Propane, Butane, Hydrogen and Natural Gas, among other gases) and smoke. This sensor detects concentrations between 300 (amount of CO found under normal conditions) and 10,000 parts per million (ppm) and operates with an operating voltage of 5 Volts and an electric current of 150 mA.

[032] Smoke sensors, based on carbon monoxide, were connected to the Arduino microcontroller and their code was uploaded. Three measurement scales were used for smoke detection: less than 300 ppm (reading <= 299); between 300 and 700 ppm (300 <leitura>=700); and greater than 700 ppm (reading>700).

[033] Next, smoke sensors based on CO2, which perform detection between 400 and 1000 ppm, and on other gases present in the burning of organic material such as ammonia gas (between 10 and 300 ppm), nitric oxide and benzene (10 and 1000 ppm), were connected to the microcontroller. All with the same pattern as the smoke sensor based on CO2 detection, with readings >= 400 ppm.

[034] After functional tests with both types of sensors connected, viewing the readings on the microcontroller's serial output, real reading tests were performed in controlled fire tests on material. Petition 870250056046, dated 02 / 07 / 2025, page 12 / 31 10 / 14 organic matter accumulated on the ground, simulating a real forest scenario, with the sensors returning the values ​​correctly.

[035] For fire / flame detection, a specific sensor was used for this purpose with detection of light waves in the infrared spectrum, emitted by fire, flames or other heat sources that have a wavelength between 760 and 1100 nm (nanometers).

[036] The flame sensor is connected to the Arduino microcontroller and has a potentiometer (trimpot) that adjusts the detection sensitivity, in a detection range of approximately 20 centimeters (cm) at 4.8 V, reaching 100 cm (1 meter) at 1 V with a response time (reaction) to detection of 15 μs (microseconds).

[037] Sensitivity to detection can be adjusted by means of the potentiometer on the sensor which regulates the digital output. It operates with an output current of 15 mA and a potential difference (ddp) of 5 V. The sensor's operating temperature is between -25 and 85 °C with an ideal detection opening angle of 60 ° (degrees).

[038] Eight (8) flame / fire modules were connected to the microcontroller, two (2) on the three sides (total of six) and two (2) on the front face of the hermetic box. The code was uploaded and all sensor readings were received at the serial output.

[039] The tests were carried out in a controlled environment, initially with fire in organic material deposited at 10 cm, 50 cm, 1 m, with a distance limit of 1.20 m and a flame column height of around 7 cm. The distance obtained for the flame height of 7 cm proved to be extremely satisfactory, with distance values ​​above that established in the datasheet, demonstrating that the sensors detected flames in the initial phase. Petition 870250056046, dated 02 / 07 / 2025, page 13 / 31 11 / 14

[040] The code also incorporates an Artificial Intelligence routine that, based on temperature, relative air humidity and the presence of smoke, generates the probability of an incident occurring at the location, thus predicting the formation of a fire and sending the message with the data to the firefighting teams in advance, preventing the fire from spreading and the environmental disaster from occurring.

[041] Data transmission was carried out using GSM (Global System for Mobile Communications) signal, as the location of the equipment, in places far from radio signals and / or wireless antennas, makes the use of GSM signal the most recommended, since mobile phone towers have satisfactory capillarity in almost all locations.

[042] The GSM module is capable of forwarding Quad-Band services and operates on frequencies in operation in Brazil and worldwide, therefore allowing SMS to be sent quickly and efficiently. It has a 3 dBi (decibel) antenna and a power supply between 3.7 and 4.2 V, with a peak current of 2 A. Due to its specific requirements, the GSM module was connected to a dedicated mobile phone (smartphone) battery with a capacity of 3.7 V and 5,000 mAh.

[043] The GSM module was therefore connected to the Arduino microcontroller and a custom code was developed so that the data, when smoke and / or fire was detected, could be sent via SMS (Short Message Service) using GPRS (General Packet Radio Services) technology.

[044] To obtain the coordinates of the equipment, a proprietary code was developed based on the concept of cell tower triangulation, which consists of measuring the distance of the device between the transmission towers and Petition 870250056046, dated 02 / 07 / 2025, page 14 / 31 12 / 14 perform calculations between the distance and coordinates of each tower, performing the calculation to obtain the geolocation of the equipment. These coordinates are stored in variables to be used for transmission, along with temperature, humidity, coordinate data, and the presence of smoke and fire, when present.

[045] After installing and testing each of the sensors, components, modules and their respective codes, with debugging, adjustments and calibration, all the codes were combined into a single algorithm (computer program), so that the equipment could operate together collecting all the data at the same time. Then adjustments were made to the code to verify the correct operating parameters and then the SMS message sending tests were carried out.

[046] To record the data collected from the various sensors, an SD Card Module was installed, which has 16 pins organized and operating in 8 pairs, which are connected to the microcontroller as follows: • Pins 1 and 2 (pair 1) ground (neutral) connected to GND; • Pins 3 and 4 (pair 2) of 3.3 V connected to the 3.3 V output; • Pins 5 and 6 (pair 3) of 5V are not connected; • Pins 7 and 8 (pair 4) of the SDCS connected to the PWM digital input 7; • pins 9 and 10 (pair 5) of MOSI connected to PWM digital input 11; • Pins 11 and 12 (pair 6) of the SCK, connected to the PWM digital input 13; • MISO pins 13 and 14 (pair 7) are connected to PWM digital input 12; and • GND pins 15 and 16 (pair 8) are not connected. Petition 870250056046, dated 02 / 07 / 2025, page 15 / 31 13 / 14

[047] The SD card module code was developed so that data is written to "csv" (Comma-Separated Values) format files, the standard spreadsheet format.

[048] The Arduino can operate at a voltage (ddp - potential difference) higher than 6 V, although it operates sufficiently well with a ddp of 5 V via connection to a USB port, which operates at that voltage, even though the microcontroller has an internal voltage and current regulator to prevent overloads and damage to the equipment.

[049] To enable the equipment to operate for extended periods, portable rechargeable Li-ion batteries with a capacity of 10,000 milliampere-hours were chosen, connected to the solar panel so that the battery could be charged during the day as long as there was sunlight. In order to avoid overcharging, since depending on solar incidence the panel can generate output voltages of up to 16 V, a voltage regulator was used between the panel and the battery.

[050] In this way, during the day the solar panel provides the energy needed for the equipment to operate, also supplying the battery, which, by powering the equipment during the night, loses its charge (discharges). In this context, the equipment can operate on for a long period in a sustainable way, with regard to energy issues.

[051] The equipment development scheme with the microcontroller, sensors and integrated modules and their respective connections was carried out by means of a layout design (sketch / drawing) as shown in Figure 2.

[052] After all tests were completed, the components were packaged in a hermetically sealed box measuring 25 x 21 x 7.7 cm, as per Petition 870250056046, dated 02 / 07 / 2025, page 16 / 31 14 / 14 Figure 4, to safeguard and protect the microcontroller, sensors, and batteries from the elements, such as sun and rain.

[053] The hermetic box offers the option of fixing it to stakes and / or attaching it to trees or other structures on the back to be at a suitable height so as not to be damaged by animals.

[054] The following hardware components were used: - Arduino Mega 2560 Microcontroller (1); - Printed circuit board for connections (13); - Temperature, pressure and humidity sensor (2); - Voltage regulator (3); - Carbon dioxide smoke sensor - CO2 (4); - Carbon monoxide smoke detector - CO (5); - Flame and fire sensor (6); - GSM module battery (7); - GSM (Global System for Mobile Communications) SIM communication module (9); - SD Card (SDCard) reader / writer module (8); - SD Flash memory storage card (9); - Power connector with USB cable (12); - Rechargeable Li-ion batteries (10); - Airtight box and - 12V solar panel (11). Petition 870250056046, dated 02 / 07 / 2025, page 17 / 31< / leitura>

Claims

1 / 2 CLAIMS 1) FIRE DETECTION AND PREDICTION EQUIPMENT IN FOREST COVER AREAS, characterized by comprising: Arduino Mega 2560 microcontroller (1); Printed circuit board for connections (13); 01 temperature, pressure and humidity sensor (2) connected to the Arduino microcontroller; Voltage regulator (3); 04 carbon dioxide (CO2) and smoke gas detection sensors connected to the Arduino microcontroller (4); 04 carbon monoxide (CO) and smoke gas detection sensors connected to the Arduino microcontroller (5); 08 flame and fire sensors adjusted by means of the potentiometer connected to the Arduino microcontroller, with an output current of 15 mA and a potential difference of 5 V and operating temperature of the sensors between -25 and 85 °C with an ideal detection opening angle of 60° (degrees) (6); GSM module battery (7); GSM (Global System for Mobile Communications) SIM communication module (9); SD card reader / writer module (SDCard) (8);SD Flash memory storage card (9); Power connector with USB cable (12); Rechargeable Li-ion batteries (10) with a capacity of 10,000 milliampere / hour connected to the solar panel; Hermetic box measuring 25 x 21 x 7.7 cm and 12V solar panel (11).; 2) FIRE DETECTION AND PREDICTION EQUIPMENT IN FOREST COVER AREAS, as per claim 1, characterized by the 8 flame / fire sensors being connected to the microcontroller, with 2 (two) connected to the three sides (total of 6) and 2 (two) connected to the front face of the hermetic box. Petition 870250056046, dated 02 / 07 / 2025, page 18 / 31 2 / 2 3) FIRE DETECTION AND PREDICTION EQUIPMENT IN FOREST COVERAGE AREAS, as per claim 1, characterized by the GSM module having a 3 dBi (decibel) antenna and a power supply between 3.7 and 4.2 V, with a peak current of 2 A, connected to a dedicated mobile phone (smartphone) battery and connected to an Arduino microcontroller. 4) FIRE DETECTION AND PREDICTION EQUIPMENT IN FOREST COVER AREAS, as per claim 1, characterized by the SD card reader / writer module having 16 pins organized and operating in 8 pairs, which are connected to the microcontroller as follows: • pins 1 and 2 (pair 1) ground (neutral) connected to GND; • pins 3 and 4 (pair 2) of 3.3 V connected to the 3.3 V output; • pins 5 and 6 (pair 3) of 5 V are not connected; • pins 7 and 8 (pair 4) of the SDCS connected to the PWM digital input 7; • pins 9 and 10 (pair 5) of MOSI connected to the PWM digital input 11; • pins 11 and 12 (pair 6) of the SCK, connected to the PWM digital input 13; • MISO pins 13 and 14 (pair 7) are connected to PWM digital input 12; and • GND pins 15 and 16 (pair 8) are not connected. 5) FIRE DETECTION AND PREDICTION EQUIPMENT IN FOREST COVER AREAS, as per claim 1, characterized by the hermetic box housing the microcontroller, sensors, modules and batteries. Petition 870250056046, dated 02 / 07 / 2025, pp. 19 / 31

Images