An adaptive train monitoring system using Internet of Things

A decade ago, the solutions to problem of providing comprehensive on board instrumentation on heavy haul trains tended to focus on ‘virtual instrumentation’. These concepts leveraged on emerging increases in capability and reductions in costs of computational power. The approach was to use a limited amount real time data that could be reliably measured (usually from a locomotive) and combine this with track databases and vehicle dynamic modelling to give progressive updates of operational states. It is now opportune to look at new possibilities in the context of the Internet-of-Things (IoT). A few issues however remain the same. One problem of the ‘virtual instrumentation’ approaches was that the systems are heavily reliant on the embedded vehicle modelling to give correct results. This means that parameters must be known ‘a priori’ and modelling must be tuned and retuned as maintenance conditions of equipment change. A solution is proposed whereby a number of IoT devices are installed throughout a train to capture operational data whilst minimising operational impact in the areas of installation and maintenance. By retaining the ‘virtual instrumentation’ system, it is argued that a limited number of motes scattered through the dynamical system can provide occasional data to the comprehensive real time virtual instrumentation. The results show that the use of a ‘virtual instrument’ as a dynamic system observer can provide reference values against which measured IoT data can be compared. A practical device concept is presented and discussed.