Evaluating Probability of Detection using an Airborne Multi-Sensor System for Search and Rescue Operations: A Human Factors Flight Test Evaluation

Canada intends to replace its aging Fixed-Wing Search and Rescue aircraft with a more advanced, sensor-equipped fleet. For this reason, there is a need to evaluate the detection performance of operators while using an advanced... [ view full abstract ]

Canada intends to replace its aging Fixed-Wing Search and Rescue aircraft with a more advanced, sensor-equipped fleet. For this reason, there is a need to evaluate the detection performance of operators while using an advanced sensor system for Search and Rescue (SAR) tasks. Under a technology demonstration program, Defence Research and Development Canada (DRDC) developed the Advanced Integrated Multi-sensor Surveillance (AIMS) system, which includes an active imager, a 3-5 micron infrared (IR) camera, a narrow and a wide field of view electro-optical (EO) camera. Two Canadian Forces Airborne Electronic Sensor Operators (AESOps) controlled the AIMS system via a station in the National Research Council Twin Otter to carry out a night-time simulated (track crawl) search pattern at the DRDC Suffield Experimental Proving Ground in Alberta. The IR camera was the primary sensor used by the AESOps with the active imager used to provide further detail. The trial range was created such that 90 objects were arrayed over the range, along flight lines. The objects comprised aluminized polyethylene blankets arranged to resemble either a crashed general aviation aircraft (e.g., a Cessna 172) or a slightly different aircraft-shaped distractor. Four possible flight paths were used to prevent the sensor operators from becoming too familiar with the arrangement of the objects on the range. For the detection task, the AIMS operator was asked to fix the gyro-stabilized sensor turret at a 30° depression angle, using an azimuth straight down the flight line. The maximum (18.2° x 13.6°) field of view was used for this detection task. Based on over 22 hours of flight data collection, the rate of detection was found to be 81.8% based on 978 total detection opportunities. This preliminary analysis is a valuable and promising start to understanding the capabilities of EO/IR sensors in a SAR role. We have demonstrated that the use of an advanced sensor can result in high probabilities of detection for idealized SAR scenarios in which high contrast targets are located in mainly dry and flat grassland. A future study will be required to examine the utility of airborne sensing systems under other conditions, such as for searches involving different altitudes, groundspeeds, terrain types, seasons, and/or object types.