Analytical Modeling for Underground Risk Assessment in Smart Cities

In the developed world, underground facilities are increasing day-by-day, as it is considered as an improved utilization of available space in smart cities. Typical facilities include underground railway lines, electricity lines, parking lots, water supply systems, sewerage network, etc. Besides its utility, these facilities also pose serious threats to citizens and property. To preempt accidental loss of precious human lives and properties, a real time monitoring system is highly desirable for conducting risk assessment on continuous basis and timely report any abnormality before its too late. In this paper, we present an analytical formulation to model system behavior for risk analysis and assessment based on various risk contributing factors. Based on proposed analytical model, we have evaluated three approximation techniques for computing final risk index: (a) simple linear approximation based on multiple linear regression analysis; (b) hierarchical fuzzy logic based technique in which related risk factors are combined in a tree like structure; and (c) hybrid approximation approach which is a combination of (a) and (b). Experimental results shows that simple linear approximation fails to accurately estimate final risk index as compared to hierarchical fuzzy logic based system which shows that the latter provides an efficient method for monitoring and forecasting critical issues in the underground facilities and may assist in maintenance efficiency as well. Estimation results based on hybrid approach fails to accurately estimate final risk index. However, hybrid scheme reveals some interesting and detailed information by performing automatic clustering based on location risk index.