As robots increasingly operate in human environments, their controllers must ensure safe and reliable behavior under real-time constraints. Although optimization-based motion planners can enforce hard safety constraints, their computational demands limit their use on complex robotic platforms. Geometric motion planning offers a real-time alternative through optimization-free, closed-form control laws with reach–avoid guarantees. However, these guarantees rely on assumptions about obstacle representations that are often violated in realistic settings. When such assumptions fail, the planner’s dynamical system may preserve invariance of the safe set but lose global attractivity, jeopardizing goal reachability.

This thesis introduces a runtime verification algorithm, called Scenario-Shield, that adapts the geometric planner’s underlying dynamical system to expand its finite-time region of attraction. The method periodically samples nearby robot configurations and performs forward simulations to approximate this region. To accelerate this process, the approach is extended by incorporating statistical uncertainty quantification: conformal prediction is used to calibrate a fast membership test for candidate states, and the scenario approach provides a principled approximation of an uncountably infinite subset of the region of attraction.

To maintain computational efficiency, the algorithm is implemented using parallel computing and integrated into a geometric motion planning toolbox with ROS. The proposed method is validated in simulation on both a holonomic ground robot and a mobile manipulator, demonstrating improved reliability over baseline geometric fabrics controllers.

The drone market has been steadily growing over the last decades, resulting in drones becoming more and more common 1. For the majority of the time, the drones provide valuable services, from inspection work and deliveries to applications in emergencies, such as search and rescue, or rapid deliveries of medical supplies. Unfortunately, because drones are available to everyone, misuse cannot be fully prevented. They can cause significant disruptions to aviation, violate privacy, or transport illegal substances, leading to financial losses, or more serious consequences. A notable incident, that caused the delay of close to 1,000 flights affecting approximately 140,000 passengers, took place in 2018 at London Gatwick Airport, where 2 drones caused the airport to shut down for over 24 hours [1]. Additionally, it has been reported by Dubai International Airport, that the estimated costs of halting airport operations due to drones resulted in losses of close to $100,000 per minute of downtime2. To counteract these hazardous situations, current anti-drone methods on the market include drone guns, quadcopters using catching systems, and radio frequency jamming systems. However, all of these methods come with their disadvantages. They either involve human interaction, have high operational costs, or cannot intercept quickly over a large area, leaving a large gap in the market for an efficient anti-drone system. The Air-Guard drone introduced in this report aims to fill this market gap, by providing a quick response time to a threat, while not compromising on neutralizing capabilities. It is a fixed-wing drone capable of autonomous visual tracking and catching unlicensed drones via a shooting net mechanism integrated with a parachute. Compared to multi-rotor drones, the Air-Guard drone concept has a longer range and higher efficiency, allowing it to be readily in the air until an unlicensed drone is detected. This is achieved thanks to its unique design inspired by bird morphology. Birds can actively morph their wings and tail surfaces to actively alter their aspect ratio, wing loading, and stability to achieve the most efficient flight configuration over a wide variety of flight profiles. Similarly, the Air-Guard drone morphs its wing and tail such that the drone can be stable while loitering, to then transition into high-g maneuvers in under a second. This allows the drone to more closely follow unlicensed drones in restricted areas and immobilize these drones in a matter of minutes autonomously. The morphing concept of the drone uses multiple actuators and elastic tendons in combination with specially designed artificial feathers to emulate bird morphology. The design also considers previously neglected areas of bird wing anatomy and incorporates bioinspired aerodynamic surfaces to delay stall and increase the maneuverability of the drone. This report is a follow-up of the midterm report, where the general configuration of the drone was set up. It aims to describe the progress on the Air-Guard project, mostly from a technical point of view, but also economic and operational aspects are covered. The UAV design is multidisciplinary and is influenced by many disciplines related to aerospace engineering. Aerodynamics, performance, structures, stability, control, and electronics considerations were combined to make the design possible. It was an iterative process, requiring careful coordination and communication between all the different departments. The result of this work translates into the design of a dual engine, 3.5 kg drone that is capable of loitering for an hour, with a maximum speed of 48 m/s. The wing span in extended configuration is 1.34 m, with a total length of 1.05 m. To neutralize the threat, the Air-Guard chases the unlicensed drone with the help of its high maneuver ability, then fires a net equipped with a parachute from the nose to capture it. The materials used are balsa wood for the fuselage and fixed-wing, while the morphing surfaces are made of aluminum and 3D-printable Celanese VECTRA A950LCP. A great emphasis was put on the sustainability aspect of the drone, which lead to an electrically powered UAV, having a structure that is 99% recyclable.