Signal Detection Theory is a theoretical framework dating back to the 1800’s. However it was not until the 1950’s before the field was thrust into mainstream psychology research initiatives by Peterson and Birdsall. Their publication “The Theory of Signal Detectability” was published in 1954 in support of University of Michigan research and was primarily focused on signal detection theory relating to electronic radar applications. The underlying objective involved manipulating criterion to either increase or decrease true and false detections of a radar signal.
The “business” of Search and Rescue (SAR), whether land or aerial based, has benefited greatly from signal detection theory concepts over the past 50 years. Our present day searchers, particularly those performing aerial searches from airplanes or helicopters, conduct searches in a systematic way in order to optimize Probability of Detection (POD). The notion of POD can be complex, however the basic fundamentals involve increasing the Coverage Area of the search area and optimizing the “detectability” of the search platforms (planes, helicopters, etc). The “detectability” is determined by calculating the Effective Sweep Width of the search platform. These are standard tabulated values derived from actual real world experimentation. Most simply stated, we define the Effective Sweep Width of the search platform (measured in yards or nautical miles typically) as the equal probability the platform will identify the search object as it will “miss” the search object. This is important for search planners, because if the environmental variables, aircraft altitude/speed, sensors employed, and search object size all affect the Effective Sweep Width, then search planners can assign search platforms optimal altitudes, track spacing, and speed to increase Probability of Detection. It’s Science!
Let’s take a look at real world experimentation using signal detection theory concepts to calculate search Effective Sweep Width (detectability). In 2002 the Potomac Management Group prepared a detailed report for the National Search and Rescue Committee (NSRC). In their research, they conducted multiple experiments using orange gloves and black trash bags as search objects in a densely wooded area in West Virginia. Their experiment generated numerous sequences documenting the distance search and rescue personnel were able to the identify or positively determine absence of the search object.
In the below chart “Orange Glove Half Sweep Width Estimator”, you can see how the experimenters determined the distance where the probability of detection and the probability of non-detection are equal. In this case of the orange glove hidden in dense wooded forest, the half effective sweep width is 16.27 meters. In the event of a similar land search, search planners would use this data to effectively space out searchers to optimize the POD and Coverage Area for a similar object and search criteria. I.e. it would not be optimal to have searchers directly next to each other covering the area nor would it be effective to space them out beyond 15 meters.
In the case of the black garbage bags (fitted in shape of human), the effective sweep width was much larger since the object was easier to spot by the searchers. I.e. they could identify it from farther away and “non-detections” also occurred further away than the orange glove sequences.
These concepts and the tabulated data which are generated are of particular importance to search planners employing aerial search platforms. Aerial platforms are often limited resources and effectively optimizing their performance and coverage is crucial to effective search operations. The below Table is an excerpt from the U.S. Coast Addendum to the National Search and Rescue Supplement. While many tables similar to this one are included in the manual, this particular table provides Search Planners a visual Sweep Width depending on variables such as size of Search Object, meteorological Visibility, and aircraft search Altitude.