Pronghorn are an iconic ungulate species endemic to western North America. Low-altitude aerial surveys are commonly used by management agencies to monitor pronghorn populations, but have disadvantages regarding safety, cost, and reliability. Low-altitude slow-speed flights leave little margin for pilot error and have resulted in injuries and even death of agency staff. A safer, more cost-effective, and verifiable monitoring method is needed. Pairing plane mounted high resolution cameras with machine learning driven automated computer vision tools has proven effective in other contexts for detecting and counting animals in natural landscapes and is well suited to the task of pronghorn monitoring. This approach would allow managers to stay out of planes and provides a verifiable visual record of detected animal. It is a challenge, however, to successfully implement such a cutting-edge method while also ensuring that it is practical for managers. Our uniquely qualified interdisciplinary team from the Wyoming Game and Fish Department, the University of Wyoming’s School of Computing and the University of Wyoming’s Department of Zoology and Physiology together have the full range of expertise needed to implement just such a tool. We plan to 1. trial a range of airplane mounted camera options and flight heights to optimally record pronghorn across vast landscapes; 2. Design and train deep learning-based computer vision models to accurately detect and count pronghorn from the collected imagery and validate model performance using management relevant metrics; 3. Document our method by releasing open source code with tutorials and producing a white paper and seminar directed at agencies in the Western Association of Fish and Wildlife Agencies. Our approach will keep managers across the western states safe while simultaneously providing a novel high-quality and cost-effective monitoring solution.