Over the past few years, there have been many advances in rail-based transport, i.e. higher speeds, increasing numbers of longer trains, shorter intervals between trains, greater axle loads, etc. Individually and together, these factors are resulting in the need for higher standards of track quality and associated infrastructure components. At the same time, railway undertakings have to minimise the expenditure on providing redundancy to cope with exceptional circumstances. Therefore, the most important performance characteristics required are greater reliability and reduced maintenance cost, while consolidating safety improvements. Turnouts are amongst the last remaining “fail hard” elements of the system railway and are exposed to severe environmental influences. There are now many situations where the failure of an individual turnout or a single switch rail can cause a total system shutdown with potentially very large losses, in terms of service reliability, operational schedules and financial consequences. Preventing such problems is therefore important for a reliable passenger and freight service. Wear is one of the most important factors affecting the reliability of railway turnouts. Consequently, a wear control method is required. Wear assessment mechanisms for railway infrastructure components often require remote monitoring of equipment condition, involving sensing of equipment states, transmission of the data collected and analysis of the signals. In any real environment, signals and measurements in condition monitoring systems can suffer from distortion and electrical noise. To be viable, the systems also require robust and reliable algorithms to detect faults. Noise could cause the system to indicate wear problems, where none exist reality. In preparing this paper, teams at the Universities of Sheffield (UK) and Castilla-La Mancha (Spain) have studied the application of remote condition monitoring (RCM) to the mechanisms used in railway turnouts and their operation. The authors put forward an algorithm to monitor wear related problems in turnout mechanisms and propose a Kalman filter for the linear discrete data-filtering problem.