R. Mills, MSc
PhD Project: Guidance and Control in the Aerial Hunting by Birds
Aerial, avian predators, such as falcons, hawks and eagles display spectacular acrobatics when hunting for their prey. Their success rate in the interception of highly maneuverable prey in the air is about 9-30%. Considering the complexity of maneuvering in the air, this is highly impressive. Of these predators, the fastest is the Peregrine falcon (Falco peregrinus), which can reach diving speeds of over 350 km/hour, while maintaining stability, control and precision in maneuvering. In this PhD project, I aim to understand the hunting behavior of Peregrine falcons and related species from a theoretical perspective. This theoretical investigation can be divided into several sub-topics:
- Guidance strategies. The interception of a maneuvering object is a challenging computational feat. The predator needs to fly towards the position that the prey is predicted to be at the same time point as the predator. From the perspective of guidance, we may ask what visuo-motion cues and guidance algorithms an avian predator uses to intercept its prey.
- Control (Aerodynamics and physical constraints). While calculating the interception, the predator needs to maintain stable flight in the air, which implies that it needs to adjust its flapping, body orientation, apply appropriate tonic muscle constriction, deform its tail and wings, and so on. The control necessary for avian flight may pose constraints on the guidance strategy applied; the predator is bounded by the aerodynamic forces that it can produce. When the guidance algorithm dictates that a force is necessary that the bird cannot produce, adherence to the guidance is compensated.
- Attack and escape strategies. Avian predators use a variety of different strategies. Of these, the most well-known is the stoop by the Peregrine falcon (a high-speed, controlled dive). Apart from the angle at which the predator attacks the prey (from above, below, or at level flight), attack strategies may be surprise (coming from behind, and approaching at high speeds from far away), long distance chases, or isolating prey from a flock. Similarly, the prey has many escape strategies: It could outmaneuver the predator by turning faster, or flee towards a safe location. Prey are often lighter than their predator, which implies that many prey can ascend faster. Thus some prey may try to escape by upwards flight. Interestingly, the flight performance of the predator and the prey, physical properties, and the guidance strategies applied determine which attack and escape strategies are most successful.
There are several modeling approaches one could adopt to study aerial hunting theoretically. Pursuit-evasion has been studied in a variety of fields, including missile guidance, optimal control of nonlinear dynamic systems and differential game theory. Between these fields, there is a lot of useful cross-fertilization. For instance, optimal guidance strategies that are found in the missile guidance literature may be fit to empirical data from animals. This way, it can be studied what control strategies these animals apply and whether they are optimal. Alternatively, biology may inspire new guidance strategies for missiles.
However, optimal strategies derived in highly idealized mathematical formulations of the pursuit-evasion problem may not take into account important biological and aerodynamic constraints, and do not portray specific attack and escape strategies found in nature. Therefore, in order to integrate all the subtopics of aerial hunting, I will build a three-dimensional agent-based simulation, in which the motion of the agents is governed by the (simplified) aerodynamics of real birds and wherein particular guidance strategies, and attack and escape strategies will be simulated. With this model, I will be able to study capture rates, flight trajectories and duration, energy expenditure and more, for many different species and strategies.
This project is supervised by prof. dr. Charlotte Hemelrijk (University of Groningen, The Netherlands) and dr. Graham Taylor (Oxford University, England) and funded by NWO open competition (dossier 823.01.017) awarded to prof. dr. Hemelrijk, entitled 'Optimisation of navigation for intercepting prey during aerial hunting by birds'.
|Last modified:||04 December 2020 10.08 a.m.|