Autonomous Targeting of Pine Processionary Moth (PPM) Nests via a Multirotor Drone

Abstract

Lebanon is distinguished by its unique ecological formation and natural charming forests that are habitats to various organisms. However, this country and other Mediterranean countries are currently facing an ever-growing threat posed by parasites, specifically known as Pine Processionary Moth (PPM), which are targeting pine trees and are exponentially increasing due to climate change. These parasites feed on pine needles causing defoliation of pine trees. While a few solutions have been proposed to reduce or eliminate the risks of these parasites, including chemical- and labor-based solutions, they remain impractical and difficult to apply. In the case of dense forests with concentrated pine trees, labor-based solutions require a large workforce and spraying pesticides in large volumes can cause severe damage to the forest ecosystem and biodiversity. Given the above constraints, the work proposed in this thesis focuses on providing a safe and automated solution that is capable of efficiently penetrating PPM nests to inject a fungi-based solution inside the nests. Since most PPM nests are located in forest areas at treetops (around 6m), a multi-rotor unmanned aerial vehicle (UAV) with vertical take-off and landing (VTOL) capabilities is utilized. The robotic system is fitted with a custom-designed injection mechanism to inject the fungi-based liquid solution inside the PPM nests. To achieve full autonomy of the injection mechanism, it is equipped with a sensing rig designed to non-visually confirm nest penetration. The main aspect that is covered in this thesis is the design, manufacturing, and assembly of a mechanically-smart injection mechanism that is able to confirm penetration of PPM nests up to a designated depth (3-6 mm) and inject an especially-developed chemical solution into their cavities, which results in one or more moths ingesting the fungi-based pesticide and then spreading the resultant infection to the entire nest's population. The experimental results on real PPM nests demonstrate that the mechanism achieved a high success rate (82%) in confirming nest penetration, in addition, to the successful and accurate penetration of the nest to a depth of 6mm, effectively meeting the target range. Future work includes performing live trials in real forests by manually operating the UAV. Once satisfactory performance is achieved in this aspect, the focus will shift to implementing autonomous PPM nest detection and targeting by integrating the various subsystems of the aerial robotic system.