Experiments to determine optimal effectiveness of drones for post-disaster search and rescue

A central and fundamental component of disaster response is the use of technology or human teams to locate and assess vulnerable people potentially trapped or injured. This set of assessment activities are generally termed “Search and Rescue” (SAR), and specifically termed “ASR Level 1” based on the International Search and Rescue Advisory Group (INSARAG) protocols. This SAR activity has traditionally been carried out by teams on foot, with search dogs or from the air using fixed wing or rotary aircraft. However, the advent of uncrewed aerial systems (UAS) has opened an opportunity for a faster response with fewer logistical overheads, potentially leading to a more effective ASR Level 1. Emergency response organisations across the world are utilising UASs for SAR activities with varying degrees of success. However, there has been no in depth research into the effectiveness of their use. The result of this is that currently exists no tested standard operating procedures.

During two sets of fieldwork, over 60 drone missions were flown, capturing data that was organised by the visibility of target, the environment, the altitude of the aircraft and the sensor type and sensor angle. The imagery collected during these missions was presented to many human observers who were given the task of determining whether each image contains a human target or not.

These observation results were analysed using a modified SAR effectiveness formula. This enabled the comparison of the effectiveness of different flight parameters leading to an evidence-based Standard Operating Procedure for drone operations in Search and Rescue post-disaster.