Guest blog by Craig Perrin
(May 30, 2023) Students from UC Davis spent 24 hours hacking for social good on May 20-21 in support of the American Wild Horse Campaign’s (AWHC) latest initiative to build scalable and robust infrastructure for wild equid fertility control programs nationwide.
Wild horses and burros are federally protected wildlife species that have a unique historical and biological role in the Western United States. Horses evolved in North America before spreading worldwide. Some bands are believed to have been reintroduced to the Americas by European colonizers and recent research indicates horses and humans worked and lived together in the Americas long before that. The descendants of these original populations roam wild across the West to this day.
The U.S. Bureau of Land Management (BLM) currently spends over $95 million every year to remove horses from public lands in the West and store them in captivity. BLM’s current removal program was shown in 2013 by the National Academy of Sciences to be making the problem worse. AWHC believes there is a better way - fertility control.
To administer fertility control programs and help keep wild horses wild, AWHC is building a new artificial intelligence (AI) app called the Wild Horse ANimal Identifier, aka Wild Horse ANI. Wild Horse ANI is a mobile app that fertility control darters will use to identify individual horses, track fertility control dosing and update horse population data directly in the field. Wild Horse ANI is needed to provide accurate dosing information to darters in the field and to support AWHC’s scientific programs that track the efficacy of fertility control. Wild Horse ANI will use the WildMe open source AI engine under the hood to power individual horse identification.
AI holds great promise to help advance fertility control by helping to ensure adequate fertility control dosing over a period of several years. However, there are significant technical challenges for wild horses, which do not have readily identifiable visible fingerprints as do species such as zebras, whales, and giraffes. Using AI for wild horses requires an extensive investment in collecting tens of thousands of horse pictures to train the AI how to identify individual horses.
To build an app to collect photos for training the AI, AWHC turned to HackDavis, the UC Davis 24-hour hackathon where over one thousand students from all over the region built apps for social good. AWHC was chosen as one of three non-profit organizations for the 2023 hackathon. AWHC’s hackathon challenge was to create an app that would make it easy for the general public to get involved in saving the wild horses.
Hackathon winners Akshat Adsule, Harsh Karia, Terry Tong, and Aarav Urgaonkar created mobile apps that not only captured horse pictures in the correct format directly from mobile phones, but made it easy to share photos to encourage others to get involved. AWHC is proud to be able to partner with these students to take their initial ideas forward to build the photo collection app in support of the overall Wild Horse ANI platform.
AWHC currently plans to launch Wild Horse ANI into production use by the end of 2023.
