This IDEA research was aimed at developing an Unmanned Aerial Vehicle (UAV)-Laser-Camera system for the railroad industry to assist railroad bridge managers to determine total transverse displacement of railroad bridges under trains. A new system was designed by integrating a UAV equipped with lasers, sensors, and other hardware while making sure its cost-effectiveness. The innovation resided in the integration of measurements from two sensors of different natures -- laser and camera. An algorithm was developed to collect various signals from the UAV sensing platform and obtain the needed total transverse displacement values. Initially, the hardware and software development were conducted in parallel. The two main sensors, laser and camera, were selected with consideration to their accuracy, range, and cost. The accuracy of the methodology was tested for obtaining the 6 DoF motion using the video collected by the camera and the computer-vision technique. The camera and laser signals were subsequently integrated to obtain the total displacement. Subsequently, laboratory experiments were conducted to validate the algorithm prior to field implementation. The new system was then field tested to demonstrate that the system could fly near bridges and measure bridge displacement and to prove its accuracy and performance while being untethered. Several test flights were conducted in the Balloon Fiesta Park at Albuquerque, New Mexico. These experiments were also aimed to identify challenges of the methodology and inform future steps to help create a robust integration between the algorithm and hardware outdoors. Due to limited access to the field because of the COVID pandemic, the railroad recommended to test the system in any type of bridge to ensure that the system was scalable for bridge environments. One local pedestrian bridge in Albuquerque of size similar to timber railroad bridges was chosen to conduct large-scale bridge experiment. Possible uncertainties and challenges of the field experiment have been investigated and addressed prior to the test. The field test provided valuable information about the uncertainties that might be experienced in real railroad bridge monitoring.  The repeatability of the results was also validated. Feedback from the railroad industry indicated its readiness for field use. The method can measure total displacement, both pseudo-static and dynamic components, under freight load which was previously not possible. Future improvements in cost and accuracy are to be expected with progress in the UAV, sensors, and laser market.

The final report is available.