Animal and aircraft collisions pose a significant risk for both animals and
humans. Thousands of animal/aircraft collisions occur annually that costs
aviation millions of dollars. Comprehensive airport wildlife monitoring and
hazard mitigation relies heavily on having accurate information regarding
diverse wildlife species habitats and land coverage, including both natural
features and human structures. Although airport biologist and personnel
attempt to mitigate these risks by deterring certain species from airports by
habitat modification (fencing, translocation, auditory and visual deterrents,
and population control), identifying areas used by wildlife and prioritizing
management actions can be difficult. Wildlife monitoring is routinely
conducted by many airports; however, frequent monitoring is sometimes
unattainable due to the amount of area needing covered, time constraints, and
lack of funding. Small Unmanned Aircraft Systems (sUAS) have recently emerged
as a potential solution for safely conducting accurate airport animal surveys
among multiple human observers. The goal of this project is to standardized
methodology and develop best practices (flying drones, capturing images, and
identifying animals without disturbance) for using sUAS in wildlife monitoring
The primary objectives of this project include:
- Calibrate sUAS flight altitude, image overlay preferences, and
selected sensors relative to representative wildlife and domestic mammal
species to standardize image collection among surveyed land coverages
representing those found within and near airport environments.
- Conduct controlled field experiments designed to assess bias in
estimating numbers (via imagery from sUAS) of available mammal species and
broad species groups common to airport environments.
- Develop imagery collection protocols according to focal species
or species groups of birds and mammals, recommended equipment, and survey
conditions for airport personnel to assist with building a test image
repository for developing automated image processing.
- Develop recommendation for image analysis tools, a deep neural
network or similar pattern classification-based system for automatic target
recognition/abundance estimation, which can eventually serve as one component
of a web application tool for airport personnel or airport contractors
interested in using UAS for wildlife monitoring.
- Make recommendations regarding sUAS-based survey design for
mammal surveys on airport properties.