Marine mammals, such as dolphins and other cetaceans, are an important indicator of ocean health. Their wellbeing relies on complex foodwebs’ functionality across the vast expanses of their habitats and vice-versa – ecosystem integrity depends on their continuous occurrence through time. Therefore, it is essential to monitor changes in their distribution and abundance and infer about apparent trends. However, due to their highly-mobile nature, this task is logistically challenging to implement. The conventional survey technique employs fixed-wing aircraft occupied by expert aerial observers who record dolphin sightings along predesigned search strips. Due to associated costs and risks, many monitoring programmes still refrain from adopting those platforms.  A collaborative study between the Apex Predator Laboratory at the Morris Kahn Marine Research Station, and the Marine Imaging Laboratory at the Department of Marine Technologies, Leon H. Charney School of Marine Sciences, investigated the feasibility of survey automation using unmanned aerial vehicles (drones). The teams conducted focal follows of striped, bottlenose and common dolphins throughout the Mediterranean Sea and produced a large dataset of annotated images captured from a range of flight altitudes. Then, they used different subsets of their data to test the performance of a state-of-the-art, deep-learning algorithm, trained for automated object detection, and establish a flight altitude which would provide for sufficient resolution on the one hand and large spatial coverage on the other – the two prerequisites of this method. Additionally, the teams compared different tracking algorithms to automate the annotation process, which is notoriously labour-intensive. The study generated the first database of annotated aerial images of dolphins available for public use.