Description:FORM 2
THE PATENTS ACT, 1970
(39 of 1970)
&
The Patent Rules, 2003
COMPLETE SPECIFICATION
(See sections 10 & rule 13)
1. TITLE OF THE INVENTION
AN IoT ENABLED AUTOMATIC CHEMICAL MANAGEMENT (ACM) SYSTEM TO ENHANCE SAFETY IN FIRECRACKER MANUFACTURING
2. APPLICANT (S)
NAME NATIONALITY ADDRESS
DIVYASAMPARK IHUB ROORKEE FOR DEVICES MATERIALS AND TECHNOLOGY FOUNDATION IN Indian Institute of Technology Roorkee, Roorkee-247667, Uttarakhand, India.
3. PREAMBLE TO THE DESCRIPTION
COMPLETE SPECIFICATION

The following specification particularly describes the invention and the manner in which it is to be performed.

FIELD OF INVENTION:
[001] The present invention relates to the field of monitoring system and method for firecracker manufacturing. The present invention in particular relates to an IoT-enabled automatic chemical management (ACM) system to enhance safety in firecracker manufacturing by monitoring and balancing chemical proportions in real time to prevent accidents.
DESCRIPTION OF THE RELATED ART:
[002] Safety of a person is primary concern in any industry especially in crackers factory. Today safety of worker is a major challenge. The health and life of the workers are critical due to the environment and its impact. The cracker factory puts the life of the workers in danger through releasing fire and smoke. The life of workers in cracker factory is unstable and day by day the pressure on them increases. The need for an innovative approach rises to safeguard the workers and to increase the productivity. The people in fireworks factory should be aware of fire accidents so that they can protect themselves before anything happens. The alarm and surveillance plays an important role in industries to communicate workers if it senses dangerous part. The area should be monitored such as temperature, toxic gas, humidity, etc. if any type of accidents sense by monitor, the immediate action will be taken to avoid the forthcoming hazards. In the project, an innovative approach is designed to sense the environmental parameter in the factory. Different types of sensors are used to design it.
[003] Reference may be made to the following:
[004] Publication No. CN2155923 relates to a sound, light, mechanic and electronic device, which can simulate firecracker setting-off and can be used repeatedly; the utility model is mainly composed of a housing, a power supply, a control circuit board, a simulated firecracker, an ignition system, a sound system, a flashing system and a mist system.
[005] Publication No. ZA202306473 relates to a chemical mixing machine for preventing the fire accidents in cracker industries due to improper mixing of chemicals. The machine comprises a container, sensing unit, an arduino, mixing unit and an outlet port. The container configured to store the chemicals for measuring the temperature, quantity and odour levels by means of the sensor unit’s temperature sensor level sensor and odour sensor. The sensors detect the deviated values to alert the workers through the arduino instruction. The proper quantity of chemical are mixed inside the mixing unit to produce the quality crackers with the help of water pipe. The present invention will reduce the accidents due to improper mixing of chemicals in the cracker and chemical industries in an effective manner.
[006] Publication No. CN202339148 relates to an electronic fire cracker, in particular relates to an air pressure explosive type electronic fire cracker.
[007] Publication No. CN111462377 discloses an intelligent access control system for firefighting, and belongs to the technical field of firefighting safety. The system comprises a face recognition module, an infrared counting module, an entrance guard switch module, a fire detection module and a master control module; when the fire detection module detects a fire, a first counting signal is sent to the infrared counting module; after the infrared counting module receives the first counting signal, people escaping from an access control unit are counted and the number of the escaping people is fed back to the master control module; the master control module determines the total number of people in the access control unit, compares the total number of the people in the access control unit with the number of the escaping people, and determines whether trapped people exist in the access control unit and the number of the trapped people.
[008] Publication No. KR101842500 relates to an intelligent fire detecting system, and more particularly, to an intelligent fire detecting system which can warn within a golden time, and can immediately react through monitoring when a fire occurs by quickly verifying and continuously warning a position of a fire detecting unit in a communication console and a fire management device according to unique ID information in each fire detecting unit when the fire occurs.
[009] Publication No. CN204331330 discloses a firework and cracker warehouse security monitoring system. The firework and cracker warehouse security monitoring system comprises a plurality of monitoring terminal systems and a monitoring computer; the monitoring terminal system comprises an ARM micro controller module, an AC-DC conversion circuit, a crystal oscillation circuit module, a rest circuit module, a hard disk storage circuit module, an RS-485 bus communication circuit module, a key operating circuit module, a digital signal processor, a digital temperature-humidity sensor, a signal preprocessing circuit, a first CCD camera, a second CCD camera, a third CCD camera, a fourth CCD camera, a smog sensor, a pressure sensor, a dust concentration sensor, a liquid crystal display and an audible and visual alarm.
[010] Publication No. CN220709755 discloses an integrated fire alarm comprising an infrared spark sensor D6, a signal processing module and a master control module, the infrared spark sensor is connected with the master control module through the signal processing module, and the master control module comprises a single-chip microcomputer module U2 and a power supply module connected with the single-chip microcomputer module U2.
[011] Publication No. KR102590048 provides a control system having a fire prediction and extinguishing function using a site-adaptive algorithm including: a fire gas data collection device collecting gas such as hydrogen chloride, air quality, carbon monoxide, and carbon dioxide generated in the event of a fire in an electrical panel or a control panel; a combined flame detection part using infrared and ultraviolet rays to detect an electric spark; a fire prediction data collection device including a non-contact infrared temperature sensor detecting a temperature change in a busbar or a power line joint part; an information collection part connected to the fire prediction data collection device by wire or wirelessly, and removing a random noise from the data collected by the fire prediction data collection device using a median filter; and a diagnosis control part receiving the data processed and outputted by the information collection part and applying each weighted value to each piece of the data to calculate a total average value to be suitable for the site status, and determining a preset phased alarm and an extinguishing step in accordance with a change of the result value thereof. Therefore, the present invention is capable of reducing damage caused by a fire by improving fire detection performance.
[012] Publication No. CN215026964 provides a PLC-based dust collector fireproof control device, which comprises a PLC, the PLC is respectively and electrically connected with a temperature sensor and a spark detector, the temperature sensor is arranged on a bag type dust collector body and an air outlet pipeline connected with the bag type dust collector body, and the spark detector is arranged on the bag type dust collector body. The spark detector is installed in an air inlet pipeline connected with the bag type dust collector body, a nitrogen pipeline is connected to the air inlet pipeline between the spark detector and the bag type dust collector body, an electromagnetic valve electrically connected with the PLC is arranged on the nitrogen pipeline, the PLC is further electrically connected with an audible and visual alarm and a control panel, and the audible and visual alarm and the control panel are installed on a control cabinet.
[013] Publication No. CN210409319 provides a spark detection and fire-fighting system for the woodworking sanding process.
[014] Publication No. CN105920761 discloses a fire-extinguishing and explosion-suppressing device for combustible dust operating spaces in the industry and trade industry. The fire-extinguishing and explosion-suppressing device comprises a spark detector for detecting and outputting a detected spark signal in a pipeline, a linear temperature sensor for detecting and outputting a detected temperature signal in the pipeline, a controller for receiving the spark signal and the temperature signal, setting an alarm threshold value of the spark signal or the temperature signal according to an internal control program, and judging whether an alarm signal is sent or not, a drive circuit for driving terminal equipment to act according to the alarm signal sent by the controller, and the terminal equipment comprising a high-pressure water pump, an intelligent feedback spray module and an explosion-proof valve arranged on the pipeline.
[015] Publication No. CN214750051 relates to the technical field of material fire resistance detection, in particular to a high polymer material fire resistance detection device which comprises a combustion detection box, a high-temperature flame injection system is arranged at one end in the combustion detection box, and a to-be-detected fire-resistant material is arranged at the other end in the combustion detection box.
[016] Publication No. CN108053603 discloses a fire monitoring alarm system of a mobile terminal. The system adopts a common sensors widely used in the market for obtaining various data such as temperature, oxygen concentration, dust concentration, delay, flame, spark and the like, and the real-time monitoring requirements of various occasions can be met.
[017] Publication No. CN203817558 discloses a detection mechanism of a fire extinguisher in a machine tool. The detection mechanism comprises a special control system for an electrical discharge machine tool, a temperature sensing probe and a fire extinguisher pressing motion execution mechanism, wherein the temperature sensing probe is connected with the special control system for the electrical discharge machine tool through a signal line; the special control system for the electrical discharge machine tool is connected with the fire extinguisher pressing motion execution mechanism through a control line; a fire distinguisher installation seat is arranged in a stand of the electrical discharge machine tool; a bottom plate is included in the fire distinguisher installation seat and is perpendicular to the gravity direction of the fire extinguisher; a weighing sensor is arranged in the upper side surface of the bottom plate and is connected with a detector through a weighing signal line; the detector is arranged in the special control system for the electrical discharge machine tool.
[018] Publication No. CN207342055 relates to the dust pelletizing system field specifically relates to a fire and explosion protection's dust pelletizing system and method. The utility model discloses a: the dust remover is connected with the gas flow tube way for remove dust to the air current, air exhaust device is connected with the dust remover for driven flow flows, spark monitor cell sets up the way in gas flow tube for the inside spark state of inspection air current, spark extinguishing means sets up the way in gas flow tube for handle the spark in the air current, the control unit, respectively with air exhaust device, spark monitor cell and spark extinguishing means signal connection, the control unit is according to the spark state control spark extinguishing means of spark monitor cell feedback and air exhaust device's operating condition.
[019] Publication No. CN116547505 relates to a device for the metered detection of fire-like phenomena in particular spark, flame or ember or hot particle phenomena in a medium flowing through or in a reservoir loaded with a medium, comprising a measuring device configured to record measurement data, said measuring device comprising a first and a second measuring unit and a third measuring unit for detecting electromagnetic radiation emitted by the fire-like phenomenon in the first or second wavelength range, optionally a third measuring unit for detecting ambient light (and optionally a sensor unit for measuring medium-specific or environment-specific measurement data. In order to reduce the risk of false alarms of such a device and to improve the detection and risk assessment of fire-like phenomena, an inspection device is provided which is adapted to inspect, at the current point in time (tA), whether a regulation criterion (A) is met on the basis of measurement data recorded by the measuring device (4) and/or stored medium-or environment-specific characteristic data, if an adjustment criterion (A) is met, the measurement sensitivity (M1, M2, M3) or the operating parameter (P1, P2, P3) of at least one of the measurement units can be appropriately adjusted for this situation using the control device.
[020] Publication No. IN202041053017 relates to firecracker disasters are continuously happening across multiple firework industries. The major reason for frequent occurrence of these fire accidents on these fireworks industries are happening due to lack of standard technique to handle the chemical mixing process.
[021] Publication No. IN202241047619 relates to lighting fireworks during festival season is a common occurrence in our country. There are some fireworks, namely firecrackers, which are not always safe. The general procedure to ignite them is using a long incense stick specifically available for that. In some incidents, the firecracker's ignition thread burns very fast and the cracker bursts before the person igniting it can get to a safe distance. Here an electric pulse spark ignitor is employed to ignite the firecracker.
[022] Publication No. IN202441022969 relates to an adjustable fire cracker bursting assistive device, comprising a cylindrical body associated with the device and accessed by a user for engaging a front portion of the body with wick of a fire cracker, a motorized iris lid is integrated on the front portion to get opened for allowing the wick to get inserted inside the body, an ultrasonic sensor integrated in the body for monitoring length of the wick inserted inside the body, an IR (infrared) transceiver is integrated on the body for receiving a wireless transmission signal from a remote integrated with an IR emitter, a heating coil arranged inside the body to get heated for igniting a flame on the wick, an artificial intelligence-based imaging unit to determine presence of a person in proximity to the to the body.
[023] Publication No. IN202311031246 relates to the examination of fireworks mishaps are just arrangements with manual prevention, mindfulness and how to keep away from mishaps during the assembling procedure of well- being matches and fireworks. Dangerous accidents frequently happen, resulting in overwhelming misfortunes of human lives and wounds to labourers. The fundamental driver of Fire mishaps is Gun Powder.
[024] Publication No. CN108205861 relates to a forest fire prevention control system. An acquisition and sensing device mounted on a drone acquires video images in a forest area to be monitored; a positioning device positions the position of the drone and generates positioning trajectory information; and a data transmission device transmits the acquired video images to a server; the server parses the video images, performs fire identification according to forest fire features, and determines a position where a fire is located according to the positioning trajectory information in the case of identifying the fire.
[025] Publication No. KR20210017137 relates to a fire detection system and a fire detection method, capable of analyzing a visible light image and a thermal image in a method of binding a single subject which is not a contour line as one, thereby detecting a fire. To this end, the real-time fire detection system and the fire detection method using the same may comprise: a photographing device acquiring a first image photographed by a visible light camera and a second image photographed by a thermal image camera; a control unit receiving the first image and the second image acquired by the photographing device by communicating with the photographing device, while controlling the photographing device; and an image analysis unit analyzing the first image and the second image in a fusion method of binding a single subject which is not a contour line as one, thereby checking whether a fire has occurred.
[026] Publication No. CN110075454 relates to a fire extinguishing robot capable of intelligently identifying flames. The fire extinguishing robot mainly comprises an AI identifying camera, a control device, a crawler belt driving device, a diesel generator device and a fire extinguishing device, wherein the generator device is connected with the crawler belt driving device and the camera, and is used for supplying power; the crawler belt driving device comprises crawler belt traveling mechanisms which are correspondingly and fixedly arranged at ports of the two sides of a chassis bracket, and each crawler belt traveling mechanism comprises a crawler belt, a driving wheel, a track roller, a tensioning device, a buffer spring, a guide wheel and bearing wheels; and an instruction for identifying the flames and then extinguishing the flames is written in by the control device, and the fire extinguishing device comprises a mechanical arm and a wind power extinguisher, wherein the mechanical arm is used for adjusting a fire extinguishing direction.
[027] Publication No. IN202421078196 relates to a system and method for detecting forest fires using advanced machine learning (ML) algorithms and image processing techniques. The system is designed to continuously monitor large forested areas in real-time using a network of imaging devices, including stationary cameras, drone-mounted cameras, and satellite imaging systems. These devices capture high-resolution visual data and thermal images, which are then transmitted to a central processing unit equipped with both graphical and central processing capabilities for analysis.
[028] Publication No. IN202421089730 relates to a drone system designed for fire detection and suppression. The system integrates advanced components such as a camera system with computer vision, infrared thermal sensors, a GPS module for navigation, and a mechanical claw to release fire extinguisher balls. Using the MAVLink communication protocol, the drone autonomously detects and locates fires, navigates to the site, and deploys fire suppression materials.
[029] Publication No. IN202541019069 relates to an AI-enhanced firefighting drone equipped with thermal imaging and first-aid deployment capabilities to improve search and rescue operations in fire emergencies. The drone utilizes a thermal vision camera to detect heat signatures through smoke and darkness, coupled with artificial intelligence for distinguishing between fire and trapped individuals. It autonomously navigates to fire locations, providing real-time situational awareness to firefighters.
[030] Publication No. IN202541026833 relates to an AI-driven firefighting drone system designed for autonomous fire detection, suppression, and monitoring across various environments. The system integrates advanced Artificial Intelligence (AI) algorithms with thermal cameras, smoke detectors, and environmental sensors to accurately locate fire outbreaks and assess their severity. Equipped with fire suppression mechanisms such as water sprayers, foam dispensers, or chemical extinguishers, the drone can deploy precise firefighting strategies based on real-time analysis.
[031] Publication No. CN115920277 relates to a prefabricated cabin type lithium battery energy storage system fire extinguishing and cooling system and a control method.
[032] Publication No. CN117482445 relates to an intelligent simple spraying fire early-stage suppression system based on the Internet of Things. The system comprises a fire type prediction system, a simple spraying device, a spraying device optimization system and a server.
[033] Publication No. CN105719444 relates to a firework and firecracker production workshop safety monitoring device comprising a workshop environment detection device and a workshop environment analytical processing device. The workshop environment detection device comprises a first MSP430 single-chip microcomputer, a first button cell, a first crystal oscillation circuit module, a first reset circuit module, a first ZigBee wireless communication module, a digital temperature and humidity sensor, a signal preprocessing circuit, a smog sensor, a pressure sensor, and a dust concentration sensor.
[034] The article entitled “Safety measures for firecrackers industry using IOT” by N. Savitha; Emerging Trends in Computing and Expert Technology; 07 November 2019 talks about the today safety is an integral part of industrial management system. Fire safety becomes the most important due to its very serious effects as a small mistake may cause severe damage. Many of the fire crackers industries manufacture firecrackers, matchbox, and printing during summer, the hot and dry climate is appropriate for manufacturing. Fireworks are the device that uses explosive, flammable material to create spectacular displays of light, noise and smoke. As in any manufacturing industry, Fireworks units also have accidents taking place in the worksite. The major cause for the fire accident are due to environmental changes and some human error. In IOT technology the fire-fighting, fire monitoring and safety management system are an important applications. So that the IOT based fire safety in the firecrackers industry is developed by placing the various sensors for monitoring the environmental parameters. All the sensor nodes are interfaced to the Arduino microcontroller and if any of the sensor node detects the abnormalities in environmental parameter the fire alert is given and once the fire is triggered the water is sprinkled over the area.
[035] The article entitled “Detection and prevention of fire accident in cracker factory using gsm” by V. Karthikeyan, Navya. R, Nidhiya Partheeban, Nithya. M, Renugha. V; International Research Journal of Engineering and Technology Volume: 08 Issue: 05; May 2021 proposes based on wireless monitoring system and effective low-cost system. The highlights of this system are easy-building up, high-reliability; powerful function and better extend ability. This system is a trendy in approach to reduce the loss in the industry and to solve the current problem. The wireless communication distance is limited in the industry and gives great protection to the industry. Detection and prevention of fire accident is an active device which is based upon arduino UNO and is used to avoid fire accident in cracker industries. The project is based on automatic fire detection. It can sense the temperature which has self-defensive ability for early prevention. It is very easy in operation as it is automatically extinguish the fire via chemical or water and a solar panel is used for power supply and it can be used for industrial purposes. GSM (Global System for Mobile Communication) are used for giving alert messages in emergency and for contact purposes.
[036] The article entitled “Control fire accident in firework industry using iot” by Mani P, Shajith Husain K, Viswa K, Shreeram R, Thanush Raja R; International Journal of Science, Engineering and Technology, 10 :3; 2022 talks about high dependable security framework and to diminish fire mishap in wafer processing plant. This undertaking centres the wellbeing of the specialists. The primary intention of this task is to protect the labourers from the fire mishaps. The IOT is cost low and powerful for remote transmission of information. This venture will lessen the responsibility of human in security upkeep. This venture is proposed in view of remote checking framework and powerful minimal expense framework. The features of this framework are simple structure up, high-unwavering quality, strong capacity and better extendibility. This framework is popular ways to deal with lessen the misfortune in the business and to tackle the ongoing issue. The remote correspondence distance is restricted in the business and gives incredible security to the business.
[037] The article entitled “Fire detection and prediction using machine learning for fireworks industry” by V. S. Sanjana Devi, Suresh Kumar B, Suresh Kumar B, Vishwa K, Vaseegaran K; International Conference on Inventive Computation Technologies (ICICT); 07 June 2024 talks about an Integrated Safety Monitoring System for the cracker industry, utilizing a combination of sensors and machine learning. Employing a DHT11 sensor for temperature and humidity, two gas sensors for hazardous gas detection, and an ESP32-CAM module for live streaming, the system provides a comprehensive view of the working environment. The K-Nearest Neighbors (KNN) algorithm analyzes historical sensor data to predict safety hazards early. This real-time system offers swift responses to emerging safety concerns, ensuring a proactive approach to accident prevention and creating a safer workplace for industry personnel.
[038] The article entitled “Spark-Break- IoT based fire control device to prevent fire accidents in firecracker manufacturing industries” talks about the spark-break- IoT based fire control device to prevent fire accidents in firecracker manufacturing industries.
[039] The article entitled “Spark detection and extinguishing system” by grupa-wolff talks about the spark detection and extinguishing system is designed to stop ignition sources from getting into areas at risk of dust explosion, such as silos or dust filters. Extremely fast and sensitive spark detectors detect ignition sources within a millisecond and then start extinguishing with a small amount of water. The extinguishing process usually lasts 5 seconds and stops automatically after which the system automatically switches to normal operation mode.
[040] The article entitled “Fire detection method in smart city environments using a deep-learning-based approach” by Kuldoshbay Avazov, Electronics, 11(1):73; December 2021 talks about a fire detector that accurately detects even small sparks and sounds an alarm within 8 s of a fire outbreak. A novel convolutional neural network was developed to detect fire regions using an enhanced You Only Look Once (YOLO) v4network. Based on the improved YOLOv4 algorithm, we adapted the network to operate on the Banana Pi M3 board using only three layers. Initially, we examined the originalYOLOv4 approach to determine the accuracy of predictions of candidate fire regions. However, the anticipated results were not observed after several experiments involving this approach to detect fire accidents. We improved the traditional YOLOv4 network by increasing the size of the training dataset based on data augmentation techniques for the real-time monitoring of fire disasters. By modifying the network structure through automatic color augmentation, reducing parameters, etc., the proposed method successfully detected and notified the incidence of disastrous fires with a high speed and accuracy in different weather environments—sunny or cloudy, day or night.
[041] The article entitled “Safety measures for firecrackers industry using iot” by N. Savitha; researchgate; January 2020 talks about the fire safety becomes the most important due to its very serious effects as a small mistake may cause severe damage. Many of the firecrackers industries manufacture firecrackers, matchbox, and printing during summer, the hot and dry climate is appropriate for manufacturing. Fireworks are the device that uses explosive, flammable material to create spectacular displays of light, noise and smoke. As in any manufacturing industry, Fireworks units also have accidents taking place in the worksite. The major cause for the fire accident are due to environmental changes and some human error. In IOT technology the fire-fighting, fire monitoring and safety management system are an important applications. So that the IOT based fire safety in the firecrackers industry is developed by placing the various sensors for monitoring the environmental parameters. All the sensor nodes are interfaced to the Arduino microcontroller and if any of the sensor node detects the abnormalities in environmental parameter the fire alert is given and once the fire is triggered the water is sprinkled over the area.
[042] The article entitled “IoT-based nfire fire alarm systems in warehouses” by nfire; 2024 talks about the warehouses are critical hubs in the supply chain, storing vast amounts of goods that, if compromised by fire, can result in significant financial losses and disruptions. Traditional fire alarm systems, while effective, often struggle to provide the level of responsiveness and data-driven insight needed to protect these large, complex spaces fully. The integration of Internet of Things (IoT) technology into NFire fire alarm systems offers a revolutionary approach, enabling real-time monitoring, enhanced detection capabilities, and automated response mechanisms tailored to the unique needs of warehouses in India.
[043] The article entitled “Smart fire detection system using IoT” by Subhankar Sarkar; Nidmjournal, Vol. 12, Issue 02, July to December 2023 talks about the fire is a chemical process between heat, oxygen and fuel, every year fires cause environmental and human loss. Fire accidents are increasing every year around the world, and India is not out of it. Every year fire causes loss of life and property damage. Controlling fire is essential to reduce all these fire-related accidents. Therefore, proper fire control planning and the development of a fire detection system are required first. If the source of the fire is known in a very short time, it is possible to repair the damage. In this case, sensor technology-based fire alarms, Smoke detectors, and heat detectors can be used. The use of IoT technology can control fires more quickly even without being present in these cases. This paper will discuss the advantages of various sensors, fire alarms, and network techniques used in fire detection using IoT.
[044] The article entitled “Wireless safety system in crackers factory and gas detection using iot” by M.Pandian, et al.; IJARIIE, Vol-9 Issue-2; 2023 talks about the fire detection using sensor networks and providing safety measures to prevent the huge fire accidents in Industries. This paper includes the design of a fire detection monitoring with Arduino based system by means of Bluetooth Module. The Arduino which controls the fire alert subjected to the temperature sensor and Smoke detector. A Fire sensor is used to detect the heat from the fire. The outputs from the Arduino are Buzzer, Water pump, LED(indicating Electricity shut down), Automatic opening and closing of door(for quick rescue purpose).All these functions are controlled by Relays which are connected with the arduino. When the system detects either the gas leakage or fire, it will immediately display the current status and alert notification on LCD display and simultaneously an alert message will be sent to the user (nearby fire station) via short message service (SMS) via Bluetooth module. Results to be obtained are, automatic fire extinguisher (water pump) will be opened in order to extinguish the fire, LED turns off which resembles the electric shut down, then the buzzer will be working for alerting the people. After fire is defected, the door opens automatically for escaping from the major issues. This will eventually allow the users to protect their lives and the properties as well from the disaster.
[045] The article entitled “Forest fire detection using CNN-RF and CNN-XGBOOST machine learning algorithms” by Umamaheswari. R; et al.; Third International Conference on Artificial Intelligence and Smart Energy (ICAIS), 02-04 February 2023 talks about the detection of forest fire should be quick and accurate as forests are the important sources to lead a vital life on earth. Detection of fire can be extremely difficult using existing methods of smoke sensors installed and they are slow and cost inefficient, so in order to avoid large scale fires, detection from visual scenes is required. In this work detection of fire in an image is done by extracting features using Deep learning algorithm and with those features as input to machine learning algorithm, a model is build with the help of different machine learning algorithms like Random Forest, Support Vector Machine, XGBoost and K-Means Clustering. Using these algorithms the data sets are classified into fire and non fire images to build the model and the test data of the data set is provided as input for getting the validation accuracy of the model. Then comparison is done among machine learning algorithms to find which algorithm provides more accuracy. To test the accuracy of the fire presence classification evaluation metrics are used in the model and find that accuracy of CNN-RF and CNN-XGBOOST are 98.53% which is greater than accuracy of CNN-SVM 97.06%.
[046] The article entitled “Grid-based urban fire prediction using extreme gradient boosting (XGBoost)” by Haeng Yeol Oh, et al.; Sensors and Materials, Vol. 34, No. 12 (2022) 4879–4890; October 24, 2022 talks about the fires in urban areas lead to enormous financial and human losses because cities have high densities of people and buildings. Although a recent advanced IoT technology improves early fire detection, it is crucial to predict fire risk to manage and prevent urban fires. We propose a method of predicting urban fires using extreme gradient boosting (XGBoost), which is based on grid-based data, to consider the characteristics of urban fires occurring in local areas. Before model training, we conducted a correlation analysis and variance inflation factor (VIF) analysis to remove variables with a strong correlation between independent variables. Furthermore, oversampling and feature selection techniques were applied to improve the model’s performance. Experimental results revealed that the overall accuracy of XGBoost was 81.25%, the F1-score was 86.43%, and the area under the curve (AUC) was 84.59%. XGBoost performed better than baseline models, such as the support vector machine (SVM) and logistic regression.
[047] The article entitled “Research on smoke fire alarm prediction model based on xgboost algorithm” by Junli Zhang, et al.; International Conference on Advances in Electrical Engineering and Computer Applications (AEECA); 2023 talks about a smog early warning model to improve the prediction accuracy. This paper takes the data collected by smoke detector as the research object, and uses machine learning algorithm to predict the smoke fire alarm. Firstly, the data diagnosis and feature engineering of smoke fire alarm data are carried out, and the feature variables are extracted by high correlation filtering method, and the feature variables are standardized. Then the data was divided into training set and test set, and the smoke fire alarm XGBoost warning model was constructed on the training set.
[048] The article entitled “Impact of fireworks industry safety measures and prevention management system on human error mitigation using a machine learning approach” by Indumathi Nallathambi, et al.; Sensors 2023, 23(9), 4365; 8 February 2023 talks about the impact of HFs that contribute to HE, which has caused FI disasters, explosions, and incidents in the past. This paper investigates why and how HEs contribute to the most severe accidents that occur while storing and using hazardous chemicals. The impact of fireworks and match industry disasters has motivated the planning of mitigation in this proposal. This analysis used machine learning (ML) and recommends an expert system (ES). There were many significant correlations between individual behaviors and the chance of HE to occur. This paper proposes an ML-based prediction model for fireworks and match work industries in Sivakasi, Tamil Nadu.
[049] In order to overcome above listed prior art, the present invention aims to provide an IoT-enabled Automatic Chemical Management (ACM) system to enhance safety in firecracker manufacturing by monitoring and balancing chemical proportions in real time to prevent accidents.
[050] The prior art predominantly uses fire detection technologies with the help of various environmental sensors which are integrated with microcontrollers to perform actions such as alarm triggers, etc. The proposed ACM machine uses industry standard sensors such as CM4-RaspberryPi, RTD100 temperature sensors, etc. Additionally, a reinforcement learning based model is deployed on the AWS cloud which assesses the probability of fire in the machine and takes actions accordingly. The live sensor data is also stored in the cloud for future training based on historical data.
OBJECTS OF THE INVENTION:
[051] The principal object of the present invention is to provide an IoT-enabled automatic chemical management (ACM) system to enhance safety in firecracker manufacturing.
[052] Another object of the present invention is to provide IoT-enabled automatic chemical management (ACM) system to monitor and balance chemical proportions in real time to prevent accidents.
[053] Yet another object of the present invention is to develop a robust, adaptive solution to mitigate fire hazards and ensure operational safety in high-risk environments.
SUMMARY OF THE INVENTION:
[054] The present invention relates to an IoT-enabled automatic chemical management (ACM) system to enhance safety in firecracker manufacturing by monitoring and balancing chemical proportions in real time to prevent accidents. Leveraging IoT sensors and machine learning models like XGBoost and reinforcement learning, the system dynamically predicts fire risks and optimizes responses such as triggering alarms or activating cooling units. The ACM employs visual, auditory, and symbolic alerts to ensure effective communication of hazards to workers, regardless of literacy levels. The system provides a robust, adaptive solution to mitigate fire hazards and ensure operational safety in high-risk environments.
[055] The invention introduces an IoT-enabled Automatic Chemical Management (ACM) system that enhances safety and efficiency in firecracker manufacturing. By integrating IoT sensors, machine learning algorithms, and reinforcement learning, the system monitors environmental factors like temperature, humidity, and pressure while dynamically adjusting chemical mixtures to prevent fire hazards. A machine learning model, such as XGBoost, predicts fire risks with high precision, while a reinforcement learning agent optimizes safety actions like triggering alarms or activating cooling units in real-time. The ACM uses visual, auditory, and symbolic alerts to effectively communicate risks to workers, including those with limited literacy. With continuous feedback and optimization, the system minimizes accidents caused by improper chemical proportions or adverse environmental conditions. This invention provides a robust, adaptive solution for fire safety in high-risk industrial settings, reducing human error and enhancing operational safety standards.
BREIF DESCRIPTION OF THE INVENTION
[056] It is to be noted, however, that the appended drawings illustrate only typical embodiments of this invention and are therefore not to be considered for limiting of its scope, for the invention may admit to other equally effective embodiments.
[057] Fig.1 shows block diagram of an IoT-enabled automatic chemical management (ACM) system.
[058] Fig.2 shows flowchart according to the present invention.
DETAILED DESCRIPTION OF THE INVENTION:
[059] The present invention provides an IoT-enabled automatic chemical management (ACM) system that enhances fire safety in firecracker manufacturing industries. This system leverages advanced IoT sensors to continuously monitor environmental parameters such as temperature, humidity, atmospheric pressure, ethanol, and hydrogen levels, which are critical to maintaining the stability of chemical mixtures used in firecrackers. The system comprises ACM, collection unit, Chemical mixture, IoT sensor temperature data, machine learning algorithm level classification of database and cooling unit.
[060] The proposed ACM machine (1) represented in figure 1, consists of three main modules. The first module is called control panel. This control module consists of the CM4 Raspberry Pi module. This is the CPU of the ACM machine and controls the SMPS power supply and VFD motor control module (2) via RS485 communication protocol. The control panel also has a temperature controller which displays the current temperature of the machine and allows us to control the temperature threshold for the cooling unit (4). Additionally, an emergency switch is provided which cuts the power supply to the motor instantly for safety.
a) automatic chemical management (ACM) (1),
b) control panel (2),
c) ACM mixture tray (3),
d) cooling unit (4)
e) IoT sensor data (5),
f) hybrid reinforcement learning based model to provide optimal action (6),
g) firecracker dashboard for chemical mixture management (7).
[061] The second module is the ACM chemical mixture tray (3). This tray is used to mix the chemicals that are used in the firecrackers. The blades which are placed inside, rotate based on the command from the control panel (2). A gearbox with an industry standard motor is fitted to the side of the machine which controls the forward and reverse rotation of the blades as well as the frequency. The tray(3) consists of holders for different sensors(5) such as temperature, humidity, pressure, ethanol and hydrogen. These sensor (5) values are saved in the cloud database and will be used to further train the model on historical data. The mixture tray (3) also consists of a jacket around it which circulates water around the machine to keep the temperature in control. The ACM machine (1) has a valve to the underside which collects the mixed chemicals.
[062] The third module is the cooling unit (4). This module consists of a submersible water pump that circulates water constantly around the ACM machine. Additionally, two cooling pads are placed to the side of the machine which are switched on based on the temperature threshold which helps in cooling the water inside.
[063] The hybrid reinforcement learning based model (6) situated in the cloud, constantly listens to the input from the raspberry Pi chip. Based on the input from the ACM machine (1), the model predicts the probability of fire based on the current state of the machine and Q-learning based reinforcement learning model predicts the best action for the machine to incorporate with an outstanding accuracy of 99.97%. If the optimal action to be taken by the machine is 0, the ACM machine performs no action as the probability of fire is negligible. Optimal action - 1, displays a pop-up on the display module with the appropriate message. Action 2 triggers an alarm on the machine which will alert the personnel nearby. Optimal action 4 represents the highest risk of fire occurrence and hence shuts down the machine instantly.
[064] The firecracker dashboard (7) enables the government officials to set the permissible limit of different chemicals components used in the firecracker mixture. Additionally, the admin will have control over the machine run and will be able to unlock/lock any machine based on the compliance. The users can enter the quantity of the chemicals being used and run the machine.
[065] Figure 2 represents the flowchart of the ACM machine. The RL model, predicts the probability of fire and the optimal action suggested by the model is carried out by the ACM model.
[066] This is an automatic chemical management (ACM) system to automate the process of chemical mixture to avoid human errors this causes major fire accidents in the firework industry. It integrates IoT sensors and an automated smart machine to mix chemicals in the firecracker industry, which employs conditional action execution. A reinforcement-learning-based fire prediction model and decision agent to predict the possibility of fire and take actions based on the reward system.
[067] Numerous modifications and adaptations of the system of the present invention will be apparent to those skilled in the art, and thus it is intended by the appended claims to cover all such modifications and adaptations which fall within the true spirit and scope of this invention.
, Claims:WE CLAIM:
1. An IoT-enabled automatic chemical management (ACM) system that enhances fire safety in firecracker manufacturing industries comprises-
a) automatic chemical management (ACM) (1),
b) control panel (2),
c) ACM mixture tray (3),
d) cooling unit (4)
e) IoT sensor data (5),
f) hybrid reinforcement learning based model to provide optimal action (6),
g) firecracker dashboard for chemical mixture management (7).
2. The IoT-enabled automatic chemical management (ACM) system, as claimed in claim 1, wherein the sensors provide real-time data to the ACM, which dynamically adjusts the chemical compositions to stay within predefined safe thresholds, thereby preventing accidents caused by chemical friction or improper mixing ratios.
3. The IoT-enabled automatic chemical management (ACM) system, as claimed in claim 1, wherein the system incorporates machine learning models like XGBoost to classify fire risk levels with accuracy 99.87%, while a reinforcement learning agent optimizes decision-making by learning the best safety actions to mitigate fire hazards.
4. The IoT-enabled automatic chemical management (ACM) system, as claimed in claim 1, wherein the actions include triggering alarms, activating fire sprinklers, or shutting down machinery and these are dynamically chosen based on the current environmental and chemical conditions.