Stochastic game network based model for disaster management in smart industry

Abstract

The smart building notion merges the pervasive sensing ability of IoT technology with the industrial infrastructure to automate decision-making during disastrous emergency. Conspicuously, a novel Stochastic Game Network (SGN) technique is proposed to reduce tangible and intangible disaster-related infrastructural losses. In particular, IoT technology is used to manage and control disasters in the SGN-based smart industry framework. Every IoT sensor functions as an individual player with predefined action sets and strategies in the proposed technique. Overall SGN of the smart industry is implemented by collaborating SGN of an individual sensor and disastrous events are analyzed using Bayesian Belief Model. The model deployed with various sensors in IoT-based industry detects disasters in advance and generates advanced warning notifications to management and control units based on game-theoretic decision-making. Simulations on 4 challenging datasets, namely Gas Sensor Array Drift, Industrial Safety and Health Analytic, Oil Spill Records, and Fire Response Times demonstrates the efficacy of a proposed model. To assess the overall effectiveness of the proposedapproach, the results are compared with other state-of-the-art decision-making techniques. Based on Gambit 14.1.1 simulations, improved findings for statistical performance parameters including Precision (93.04%), Specificity (92.35%), and Sensitivity (93.36%) were registered. Also, the mathematical analysis was performed to evaluate the effectiveness of the presented methodology