Predictive Network Modelling with Live Sensor Data

Advanced fibre-optic based sensing systems are fast growing and replacing conventional sensors inmany industries due to their reliable measurements, fast response, and intrinsically safe operations. Thedeployment of any sensor system however remains a potentially expense and disruptive activity and maybe limited in areas with difficult or unsafe access, or without communication systems. A predictive modelling tool that accepts live data and outputs both real and predictive estimationsbetween sensors may have great application by limiting the number of sensors required, or coveringareas unable to be monitored by sensors. A transient based simulation model incorporating real-timeactual and simulated data can be merged to display predicted conditions through a mine or tunnellingsystem at any time. “Ventsim Visual” software is used as a graphical interface to demonstrate theeffectiveness of this approach.