Inspired by talon of a predator bird, this paper presents a quadrotor with two 2- Degree of Freedoms (DOFs) compliant manipulators which could mimic bird perching and grasping. The symmetric configuration of the manipulators causes minimum shift in center of gravity and a minimum disturbance to the angular momentum of the platform during grasping and perching maneuvers. Thus, the dynamics of the manipulator is independent to that of the platform. Moreover, a compliant end-effector is introduced to decouple the dynamics of the unmanned aerial vehicles from the force interaction with the environment or target objects. Therefore, aerial manipulation problem is significantly simplified due to the minimum amount of disturbance among components. In addition, the manipulators could function as the landing gear, which allows larger work envelope, weight saving and less landing impact. It also has the potential to achieve a bird-like “perch and watch” to increase the endurance of unmanned aerial vehicles in missions that involve extended endurance.

Detecting unexpected objects (anomalies) in real time has great potential for monitoring, managing, and protecting the environment. Hyperspectral line-scan cameras are a low-cost solution that enhance confidence in anomaly detection over RGB and multispectral imagery. However, existing line-scan algorithms are too slow when using small computers (e.g. those onboard a drone or small satellite), do not adapt to changing scenery, or lack robustness against geometric distortions. This paper introduces the Exponentially moving RX algorithm (ERX) to address these issues, and compares it with four existing RX-based anomaly detection methods for hyperspectral line scanning. Three large and more complex datasets are also introduced to better assess the practical challenges when using line-scan cameras (two hyperspectral and one multispectral). ERX is evaluated using a Jetson Xavier NX edge computing module (6-core CPU, 8GB RAM, 20W power draw), achieving the best combination of speed and detection performance. ERX was 9 times faster than the next-best algorithm on the dataset with the highest number of bands (108 band), with an average speed of 561 lines per second on the Jetson. It achieved a 29.3% AUC improvement over the next-best algorithm on the most challenging dataset, while showing greater adaptability through consistently high AUC scores regardless of the camera’s starting location. ERX performed robustly across all datasets, achieving an AUC of 0.941 on a drone-collected hyperspectral line scan dataset without geometric corrections (a 16.9% improvement over existing algorithms). This work enables future research on the detection of anomalous objects in real time, adaptive and automatic threshold selection, and real-time field tests. The datasets and the Python code are openly available at https://github.com/WiseGamgee/HyperAD, promoting accessibility and future work.

This article presents a cost-effective camera network system that employs neural network-based object detection and stereo vision to assist a pan–tilt–zoom camera in imaging fast, erratically moving small aerial targets. Compared to traditional radar systems, this approach offers advantages in supporting real-time target differentiation and ease of deployment. Based on the principle of knowledge distillation, a novel data augmentation method is proposed to coordinate the latest open-source pre-trained large models in semantic segmentation, text generation, and image generation tasks to train a BicycleGAN for image enhancement. The resulting dataset is tested on various model structures and backbone sizes of two mainstream object detection frameworks, Ultralytics’ YOLO and MMDetection. Additionally, the algorithm implements and compares two popular object trackers, Bot-SORT and ByteTrack. The experimental proof-of-concept deploys the YOLOv8n model, which achieves an average precision of 82.2% and an inference time of 0.6 ms. Alternatively, the YOLO11x model maximises average precision at 86.7% while maintaining an inference time of 9.3 ms without bottlenecking subsequent processes. Stereo vision achieves accuracy within a median error of 90 mm following a drone flying over 1 m/s in an 8 m $\times$ 4 m area of interest. Stable single-object tracking with the PTZ camera is successful at 15 fps with an accuracy of 92.58%.

This paper presents an implementation framework to perform a vision-guided, bat-like inverted perching maneuver with a bi-directional thrust quadrotor platform. The framework consists of several distinct modules (guidance, motion planning, control, state estimation) that can be easily be individually customized in the future to meet specific research requirements. The main contribution of this paper lies in the whole framework pipeline with a modular structure developed for implementing a generalized framework for an agile quadrotor to achieve inverted perching. A computationally-light guidance module has been developed as an example to demonstrate the capability while being independent of accurate pre-known target information, and does not require the state estimation of the quadrotor to be provided by an external motion capture system as in our previous work. A motion planning module based on an optimization method has been introduced to generate a two-stage inverted perching trajectory aiming at minimizing altitude loss during the half-flip maneuver. A control module has been developed to enable a bi-directional quadrotor to fly in both upright and inverted states and closely follow the intended trajectory. The compensation strategy used in the control module is key to minimizing the transition time between the upright and inverted states. Finally, an experimental flight platform has been developed to demonstrate the capabilities of the framework. During testing, the proposed framework has achieved an 80 % success rate. To the best of our knowledge, this paper presents the first time a quadrotor has achieved the inverted perching maneuver using onboard vision guidance.

Pest birds are a salient problem in agriculture all around the world due to the damage they can cause to commercial or high-value crops. Recent advancements in Unmanned Aerial Vehicles (UAVs) have motivated the use of drones in pest bird deterrence, with promising success already being demonstrated over traditional bird control techniques. This paper presents a novel bird deterrence solution in the form of tethered UAVs, which are attached and arranged in a grid-like fashion across a vineyard property. This strategy aims to bypass the power and endurance limitations of untethered drones while still utilising their dynamism and scaring potential. A simulation model has been designed and developed to assess the feasibility of different UAV arrangements, configurations, and strategies against expected behavioural responses of incoming bird flocks, despite operational and spatial constraints imposed by a tether. Attempts at quantifying bird persistence and relative effort following UAV-induced deterrence are also introduced through a novel bird energy expenditure model. This aims to serve as a proxy for selecting control techniques that reduce future foraging missions. The simulation model successfully isolated candidate configurations, which were able to deter both single and multiple incoming bird flocks using a centralised multi-UAV control strategy. Overall, this study indicates that a grid of autonomous ground-tethered UAV platforms is viable as a bird deterrence solution in agriculture, a novel solution not seen nor dealt with elsewhere to the authors’ knowledge.