Multivariate inspection of German steel civil infrastructure using autonomous UAS

Abstract

Aging civil infrastructure and increasing traffic loads lead to deteriorating assets and the need for regular manual inspections, which generally entail at least partial closure of the asset and operational downtime. These costly inspections can be overcome by employing autonomous unmanned aircraft systems (UAS) to carry out the required inspections. Utilizing a variety of sensor configurations, AI-based analysis, and data fusion through a web-based interface, more objective comparisons between infrastructure assets can be obtained quickly. Within the dtec.bwMISDRO project the Norderelbbrücke of the German Autobahn, Freihafenelbbrücke of the Hamburg Port Authority, a railway bridge over the river Oste from Deutsche Bahn and a barrage over the river Weser from the German Federal Waterways and Shipping Authority are to be inspected via UAS with visual cameras, lidar, hyperspectral imaging and further sensors to detect typical steel defects. In this paper the planned non-destructive testing methods for automatic defect detection, the automatic path planning, and the design of the UAS are presented.