Content area
Full text
Public safety officials often institute road closures in the event of major vehicle crashes in order to provide a safe environment in which to document the crash scene. However, interstate closures result in unexpected congestion that may cause secondary crashes. A study of interstate crashes in Indiana, USA found that of the approximately 15,000 crashes in 2014, 3,448 occurred during congested conditions.1 The study found that the crash rate during congested conditions was approximately 24 times greater than during uncongested conditions.
Incident management teams train first responders to reduce clearance times.2 Since 2009, Indiana has promoted the use of photogrammetry to reduce crash investigation times. Unmanned Aircraft System (UAS) can likely further reduce the time required to accurately document crash scene geometry.
The objective of this paper is to evaluate the application of a consumer-grade UAS, costing approximately $2000, to construct ortho-rectified images of a staged crash scene and assess the spatial accuracy of the ortho-rectified images. A detailed field procedure for scene preparation, establishment of control, and data acquisition is provided. A simplified mission design table and concepts are provided for use by public safety officials.
Literature Review
As incident duration increases, the likelihood of a secondary incident also increases, which, in turn, increases the incident duration and likelihood of additional crashes. There has been a nationwide effort in recent years to improve and communicate best practices for incident clearance. Total stations have been used in incident management for decades. One early study found that the use of total stations saved on average 51 minutes and was safer for responders compared to traditional methods.3 However, at the time of this study (1992), total stations could be cost prohibitive and result accuracy was affected by weather conditions. Another study found that the number of measurements taken doubled and the time to collect data decreased by 33 percent when using total stations.4 More recently, Ardestani et. al. proposed a prototype UAV system and developed 3D models of a staged incident using video and 2D images.5 Liu et. al. also used 2D imagery of a staged crash scene collected from a UAV to generate a 3D model.6 Examples of ortho-rectification of photos obtained from UAS for crash scene documentation were published by He et. al.7-10
Staged Crash...