ABSTRACT

In late 2015, AURIZON Network engaged Australia’s leading research commercialisation company, the University of Queensland’s UniQuest, and Imagus Technology, to develop an effective and low cost video object recognition algorithm for real-time detection of the overhead electrified contact wire used with narrow gauge (1067 mm) heavy haulage electric locomotive pantographs. Control of the lateral offsets (or stagger) of the contact wire in relation to the track structure alignment is vital for electrified railway networks. Out of tolerance staggers can result in dewirements that can cause severe disruption and significant costs to network providers. This automatic Video Pantograph Electrified Wire initiative (VIPER) will support Network’s Triple AAA programme with a cost effective high frequency assessment of the stagger of its Overhead Line Electrified (OHLE) contact wire, providing baseline information, trends and exception reporting to enable predictive maintenance and prescriptive intervention by qualified analytic managers.