The data used to analyse and predict the degradation of tram tracks is divided into three main categories including: inspection data, load data and repair data. These three main categories are explained in the following sections.
The degradation of rail track geometry is usually quantified in terms of track geometry defects and rail surface faults. Track geometry defects include: longitudinal level defects, horizontal alignment defects, gauge deviations and track twist (Quiroga & Schnieder, 2013). According to the literature, gauge deviation has been considered as a popular indicator of track geometry defect (El-sibaie & Zhang, 2004; Shafahi & Hakhamaneshi, 2009; Sadeghi, 2010). Consistent with the literature, gauge deviation can be used to measure the track geometry defects. Gauge deviation is provided by Yarra Trams inspections of tram tracks.
The inspection data used in this study has been collected for Melbourne tram network between December 2009 and December 2013 over various segment types of several routes. The segment types include crossover, curve, H-crossing, and straight sections. According to the literature (Yousefikia, Moridpour, Setunge, & Mazloumi, 2014; Moridpour & Hesami, 2015), despite forming almost 10% of total length of routes under study, curves are more likely to experience exceeding the failure limits, in comparison to other segment types such as the H-crossing, Crossover and Straight segments. Therefore, for the purposes of this analysis, the focus is on curve segments. The inspection data includes the gauge values at curve segments at every 10 cm.