Dr. Robert Koester's Lost Person Behavior and How to Contribute to ISRID
Hundreds of thousands of people go missing each year, with their safe return relying on fast and coordinated search and rescue operations. Robin was joined by Robert Koester, developer of the International Search and Rescue Incident Database (ISRID) and author of 'Lost Person Behavior' to chat about what ISRID is, how teams use it, how D4H can collect data, and how to send that data to ISRID to contribute. Robert also discussed the second edition of his book 'Lost Person Behavior' which helps rescuers look in the right place to find lost subjects faster.
What is Lost Person Behavior
Lost Person Behavior is a book written by Dr. Robert J. Koester with a simple purpose - to help rescuers look in the right place to find lost subjects faster. Based upon a landmark study, the book uses clusters of behavior patterns to predict where a lost person might be found. It includes subject categories, behavioral profiles, up-to-date statistics, suggested initial tasks, and specialized investigative questions. ‘Lost Person Behavior’ is also available as an app for Android and Apple which contains all the statistical data that’s in the book and provides guidance on where a missing person might be, what to ask, and where to look.
Not everyone who is missing is lost, but everybody who is lost is missing.
Robert Koester, developer of the International Search and Rescue Incident Database (ISRID) and Author of ‘Lost Person Behavior’
What is ISRID?
ISRID stands for the International Search and Rescue Incident Database. The database was developed by Dr. Robert Koester as part of a USDA grant which allowed the collection and analysis of SAR statistics from around the world. So far, over 150,000 SAR incidents have been collected as part of the project. The data collected is used to generate models and probability statistics of how a lost person is expected to behave and where they are likely to go based on their profile. ISRID can be used as a starting place when you know nothing else about a lost individual, it can be used to place that person into a category, for example, missing hiker, or missing hunter to then predict their likely behaviors and movements.
The Lost Person Behavior Models Explained
The Displacement Model
The Displacement Model, also known as the ring model is the most common, and it uses range rings. Used for example when searching for a lost child aged 1-3, using the Displacement Model, the smallest ring is a 25 percentile, 25% of missing children between the ages of 1-3 are found within a 200-meter radius, 50% are found within 300 meters, and 75% are found within 600 meters 95% go as far as 4.5 kilometers.
The Dispersion Model
The Dispersion Model is for when rescuers are given a direction of travel. It predicts how close to that line of direction a lost person is going to be found. For example, the Dispersion Model measures how far a lost person is likely to wander, and in what direction. It is a profile of a person who can walk a straight line and has a purpose in their direction, versus somebody who doesn’t. When supporting finding a person suffering from dementia, Dispersion Model research has shown that patients with dementia have quite a tight dispersion angle.
Probability Density and Linear Features in Search and Rescue
The concept of probability density, and linear features, is where most lost people are found, for example, a lot of people are found on roads or trails, other linear features include fences and railroads, which is often where a lost person will wind up.
The Elevation Model
The Elevation model is the probability of a lost person going uphill versus downhill. Historically people were prone to going downhill, however with the use of cell phones the trend has changed that people are brought uphill for signal.
The Track Offset Model
The Track Offset Model is particularly valuable in the search for some of the most vulnerable people, such as children and dementia patients, those who may not be found on a linear feature. Track Offset helps identify how far away they may be from that linear feature. For example, of people suffering from dementia, typically 50% of people not found on a road or trail are found only 15 meters from the linear feature.
The Watershed Model
The Watershed Model is one of the most technical models. Where water flows into a common source such as a river, the Watershed Model predicts how a lost person will behave in relation to the water source. The ISRID study has shown that most lost people are unlikely to switch from one watershed to the next., 50% of lost people are found in the same watershed.
The History of ISRID and Lost Person Behavior
Dr. Robert Koester began gathering data in 2002, working with the US Department of Agriculture, with a 6-month grant and an additional 2 years of analyzing the data, he collated an international database from 43 different sources and analyzed 50,000 data submissions. It is with this data he wrote his book ‘Lost Person Behavior’. In 2013, with additional funding received from the Department of Homeland Security Science and Technology Directorate, he was able to collect an additional 100,000 new international cases which will support the Second Edition of ‘Lost Person Behavior’.
Upcoming additions to ISRID
Dr. Robert Koester is actively continuing to work on gathering more data and subject categories for ISRID and the Second Edition of ‘Lost Person Behavior’. An example of a new subject category being added is Scenario Categories. Instead of looking at what a lost subject was doing at the time of getting lost, it looks at why they went missing, for example, a medical event such as a hiker with a diabetic crisis, this would fall under the medical category.
How to Use D4H to Collect Lost Person Behavior Data and Submit it to ISRID
A critical component of Dr. Koester’s work relies on the assistance of search and rescue personnel and them sending him their own collected data Teams can contribute data via email to firstname.lastname@example.org. D4H also has a Lost Person Behavior module that was built to the silver standard for ISRID data collection. D4H Users can easily add a new incident, name the incident, and set the location using the map function, add in all their time stamps, input weather data, move it to the Lost Person Tab, and access the ISRID compatibility form. Using the initial planning point, users can work through the ISRID categories and create a compatible silver report.At the year’s end, you can select the date range for the incidents you wish to report and email the spreadsheet with a brief description of the incidents you wish to report to Robert at email@example.com. You can find more information on exporting Lost Person Behavior data from D4H to submit to ISRID in this help center article.