Is TIR Enough to find Avalanche Victims in the Colorado Rockies?

Abstract
The use of UAVs for personal and commercial data collection has grown tremendously in the last decade. Images from a UAV can be more convenient, cost-effective, and maintain high resolution making them attractive to many users. The application chosen for this project is search and rescue efforts in snowpack using thermal infrared imaging. Victims of an avalanche have a small window of time for survival that depends entirely on the promptness of search and rescue. Utilizing methods of GIS, remote sensing and a UAV, this project aims to provide a solution for a quick and reliable search and rescue mission.

Introduction
Unmanned aerial vehicles have been around since the mid-1800s, even before manned aerial vehicles. Hot air balloons and kites were some of the first platforms that were utilized by the military to perform aerial bombardment and reconnaissance missions (Watts et al. 2012, 1672). Some early experiments of airborne photography even involved strapping small cameras to pigeons, although not really considered an aerial vehicle (Colomina and Molina 2014, 81). Since then, technology and breadth of application have progressed tremendously. Modern day UAVs come in all shapes, sizes, and names. A UAV1 can also be considered UAS (which refers to all the components of a remote control flight, including the UAV), an RPA, or a drone (Wikipedia 2014). The word drone tends to draw a negative connotation as the term is widely used in military context and is perceived as being harmful. Besides military applications, data collection by UAV in conjunction with GIS techniques and remote sensing have been making waves in the civilian and commercial sectors over the past decade. Improvements over the years of UAVs as well as technology in all software and hardware components have made unmanned aerial data collection a realistic endeavor.

The FAA closely monitors UAV operations in the NAS for safety and security reasons. In some areas, UAVs are strictly prohibited. The FAA has three categories of unmanned aircraft operations. They are Civil, Public, and Model Aircraft operations. Civil UAV operations require a Special Airworthiness Certificate for flight authorization. This certificate does not permit operations for compensation, but permits operations intended for research and improvements. As of recently, this certificate can be waived with a Section 333 waiver from the FAA. Public UAV operations include any government related missions that require civil airspace. These operations require a Certificate of Authorization from the FAA for flight approval. Lastly, Model Aircraft operations refer to recreational use of the NAS and does not require special permission from the FAA. In order to operate under this category, aircrafts are recommended to fly below a ceiling of 400 feet, away from other air traffic, and always stay within the pilot’s sight (FAA 2014). Since the FAA Modernization and Reform Act of 2012, UAV enthusiasts and researchers have been waiting for new laws to emerge that might allow more people to utilize the NAS without special allowance from the FAA. The FAA has been directed to create new laws that would bring UAVs into the NAS by 2015 (Wing et al. 2013, 342).

Advances in technology of cameras and sensors that can be mounted to a UAV have just recently boosted civilian interest. Cameras have been formatted to minimize weight while retaining resolution and functionality. Users are not only utilizing visible light for data collection, but also the infrared bands of the light spectrum which opens the doors for all kinds of aerial data collection and analysis.

There are many ways that UAVs can be paired with TIR sensors to solve problems. This project focuses on avalanche search and rescue missions and how TIR sensors on board a UAV could be helpful in locating recently buried victims. The research question this project addresses is: Using a UAV with a mounted thermal infrared sensor, is it possible to detect body heat under snow? If so, how deep? Despite all attempts with several different variables, longwave infrared heat detection under even a thin layer of snow was unsuccessful. This paper outlines the pilot project that tested this research question and discusses the technical aspects of the project.

Click here for the complete whitepaper.