Design and application of the Fuzzy Logic Optimization Technique in early fire detection for low-cost fire detection systems using the IoT based platform. A case study of local Urban markets in East African (EA)

Abstract

East Africa's local markets, such as Owino, Uganda; Gikomba, Kenya; and Gisozi, Rwanda, have faced a high prevalence of uncontrolled fire accidents. Electrical circuits, arson, fuel spillage, and charcoal stoves are among the major causes of fires. Currently, fire departments have not developed any contingency plans for the management and control of fire accidents. And there are no early warning systems to monitor and control fires before becoming uncontrollable. Instead, human patrol methods are used to monitor these fire incidents, which are ineffective, insufficient, and obsolete. However, if these fires are not managed and controlled properly, market occupants and the nearby community face a high risk of severe property damage and loss of life. According to studies on unit smoke, flame detectors are unreliable, making them highly prone to false alerts. Additionally, satellite systems in developing countries are highly expensive to purchase and operate and cause unnecessary delays because of their lengthy scanning cycles for fire detection. The study utilized the MATLAB 2018a fuzzy toolbox and Arduino (IDE) software to design and develop a hardware solution for a low-cost IoT-enabled fire detection system for local markets, yielding three major contributions: A fuzzy-based detection model for early fire detection to aid public safety and control in local markets is proposed. Using MATLAB, six input parameters, i.e., temperature, humidity, flame, CO, CO2, and O2, are used vis-à-vis EFIP. Results show that, the obtained solution achieved an EFIP output accuracy of 95.83% using fuzzy inference rules. Hence, the study helps firefighting professionals and fire rescue departments understand the significance of fire control systems, with the primary goal of minimizing related fire risks and damage through the use of early warning systems. Secondly, an Interval Type-2 (IT2) Tang Sugeno Kang (TSK) fuzzy model for intelligent fire intensity detection algorithms with decision-making in low-power devices is proposed using a free open-source IT2 MATLAB toolbox. Results show that the model exhibited an accuracy rate of 98.2%, with MAE = 1.3010, MSE = 1.6938, and RMSE = 1.3015. Thus, this study serves as the basis for the development of in-built low-power fire detection systems that are economical, replicable, and quickly deployable for usage by the firefighting personnel in LDCs, with the major objective of safeguarding extremely dangerous urban marketplaces and public gazette areas from fire accidents. Thirdly, a hardware solution for a low-cost IoT-enabled fire detection system using fuzzy application methods is presented. Using the Arduino IDE, results demonstrated an accuracy of 91%, evaluated using the confusion matrix model. Hence, the study is significant for firefighting professionals as a foundation for a new approach to low-cost fire detection to respond to them urgently and quickly through providing early warning notifications and calling for quick action in case of a fire accident. In conclusion, because of the extensive fire outbreaks and a lack of effective fire control methods, the study focused on local markets particularly, in Uganda. Existing manual techniques lead to severe damage from unnecessary delays in calling police agencies. Hence, the study proposed a hardwarebased solution for a low-cost fire detection system using fuzzy reasoning to enhance early warning notifications, improve decision-making, and minimize false alerts. Future work encourages rigorous fire safety risk assessments, fire safety management, and appropriate control and suppression procedures for the markets by investigating the level of fire dangers, associated risks, and threats and evaluating the damage and expenses realized. This contributes to accurate fire risk assessment models, appropriate fire safety laws, best practices in fire safety management, and control techniques