By Shane Brunker - Technical Director at NM Group A lot of remote sensed data has been captured on infrastructure networks. A lot. It’s a range of data types for a range of purposes – LiDAR for vegetation management, track based imagery for rail infrastructure assessment, high resolution video on pipelines, multispectral satellite imaging for wide area change detection, and so on. It’s probably the LiDAR (and to a lesser extent the imagery) that interests me most as it lends itself better to cost effective computational approaches. Why do I describe it like cash in the bank? Well in the case of LiDAR and imagery being collected for the thermal rating of a transmission line, the output is generally an engineering model in PLS-CADD and a report describing capacity of the line. That’s about it. Separately, it’s quite likely that the vegetation management group have been flying helicopters to scope that line for clearance violations. Someone else might have been out there on the ground with a GPS picking out tower positions to update the GIS. You might also have the land management folks trying to determine where fences have moved, dirt has been dumped and sheds have been extended! A lot of the information being chased through these activities is already sitting there locked away in that LiDAR and image dataset – ready to be extracted.
Above, a tree objectification and risk model for a rail line based on data collected for asset management
Of course there are organizational exceptions to the scenario above – but rarely do I see data collection on networks being viewed in a holistic fashion, looking across the full scope of users and applications trying to see how the data and information can help the wider group. The ability to detect change, understand risk and improve operational efficiency is there for the taking through these information rich datasets – even more so if there are a few datasets on the same area, collected over a 5, 10 or 15 year period. I’d suggest go back and see if you actually have the data handy and if the archive it is managed by your vendor(s), check the formats and make sure it’s not in something proprietary or difficult to access. Explore the coverage and resolution and the times of year it was captured and analyzed. Engage the asset management, environmental and engineering groups among others to understand their challenges and look to industry to see what tools and outputs are available to exploit these existing datasets. The best thing about this exercise? Next time you go to market to collect something new you’ll find yourself exploiting it better from the start!