Unmanned aerial system (UAS) for railway inspection

This study considers the implementation of multi-rotor unmanned aerial systems (UAS) as railway inspection tools and aims to address the issues associated with two main categories of railway inspections - track inspection in remote areas and inspection in confined railway tunnels and culverts. For track inspection in remote areas, the UAS can dock with a moving train heading to such an area. With the help of Global Navigation Satellite Systems (GNSS), the UAS takes off and follows the rail as well as inspect the target area. The UAS then lands on the rail track and capture images while driving on the track, and the acquired images are used in reconstruction of the three-dimensional railhead profile using the Multi-View Stereo (MVS) algorithm. For inspection in confined spaces, the UAS traverses through a tunnel in a semi-autonomous fashion. Using on-board Light Detection and Ranging (LiDAR) scanner, the system is able to measure the cross sectional shape of the culvert in real time. Extra sensor data are used to measure the travel distance, then they can be further processed to obtain three-dimensional point clouds of the tunnel’s structure. These systems should be able to reduce the heavy reliance on manual inspections and provide a safer, more flexible and safer way to assist the existing inspection regimes at a lower cost.

Media Info

  • Pagination: 9p. ; PDF
  • Monograph Title: AusRAIL 2018, Rail for a better future, 27-28 November 2018, Canberra, Australian Capital Territory

Subject/Index Terms

Filing Info

  • Accession Number: 01708040
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: ATRI
  • Created Date: Jun 18 2019 4:58PM