Journal of Field Robotics最新文献

Autonomous underwater vehicles (AUVs) are highly nonlinear, coupled, uncertain, and time-varying mechatronic systems that inevitably suffer from uncertainties and environmental disturbances. This study presents an intelligent hybrid fractional-order fast terminal sliding mode controller that utilizes the positive aspects of a model-free control approach, designed to enhance the tracking control of AUVs. Using a nonlinear fractional-order fast terminal sliding manifold, the proposed control approach integrates intelligent hybrid sliding mode control with fractional calculus to guarantee finite-time convergence of system states and provide explicit settling time estimates. The nonlinear dynamics of the AUVs is modeled using radial basis function neural networks, while bound on uncertainties, external disturbances, and the reconstruction errors are accommodated by the adaptive compensator. By using a fast terminal-type sliding mode reaching law, the controller exhibits enhanced transient response, resulting in robustness and finite-time convergence of tracking errors. Using fractional-order Barbalat's lemma and the Lyapunov technique, the stability of the control scheme is validated. The effectiveness of the proposed control scheme is validated by a numerical simulation study, which also shows enhanced trajectory tracking performance for AUVs over existing control schemes. This hybrid technique addresses the complicated nature of AUV dynamics in unpredictable circumstances by utilizing the advantages of model-free intelligent control and fractional calculus.

Currently, the majority of robots equipped with visual-based simultaneous mapping and localization (SLAM) systems exhibit good performance in static environments. However, practical scenarios often present dynamic objects, rendering the environment less than entirely “static.” Diverse dynamic objects within the environment pose substantial challenges to the precision of visual SLAM system. To address this challenge, we propose a real-time visual inertial SLAM system that extensively leverages objects within the environment. First, we reject regions corresponding to dynamic objects. Following this, geometric constraints are applied within the stationary object regions to elaborate the mask of static areas, thereby facilitating the extraction of more stable feature points. Second, static landmarks are constructed based on the static regions. A spatiotemporal factor graph is then created by combining the temporal information from the Inertial Measurement Unit (IMU) with the semantic information from the static landmarks. Finally, we perform a diverse set of validation experiments on the proposed system, encompassing challenging scenarios from publicly available benchmarks and the real world. Within these experimental scenarios, we compare with state-of-the-art approaches. More specifically, our system achieved a more than 40% accuracy improvement over baseline method in these data sets. The results demonstrate that our proposed method exhibits outstanding robustness and accuracy not only in complex dynamic environments but also in static environments.

The multimodal land-air aircraft combines the advantages of traditional drones and ground unmanned equipment. It can cross obstacles on the ground, such as lakes and mountains, and fly quickly in the air, reaching a wider range. It can also switch to an energy-saving mode based on the characteristics of the surrounding environment and mission requirements, reducing energy consumption and noise while increasing endurance. Based on the idea of reusing the same structure, we have designed a multi-mode agile land-air aircraft, abbreviated as ALAA. ALAA has eight actuators, and it combines propellers, wheels, and gearboxes in different ways to achieve multiple modes of locomotion on the ground and in the air: flight mode, driving mode, and upright mode. In propeller-assisted driving mode, it can climb slopes up to 50°. It can also combine driving and upright modes, demonstrating strong obstacle-crossing capabilities. In addition, ALAA reuses the same components, simplifying the transition between flight and ground movement without the need for deformation, thus enabling fast and rational mode transition suitable for complex environments. Ground modes can extend the endurance time of ALAA, and experimental results show that ALAA can operate 21 times longer than an aerial only system. This paper presents the overall design and mechanical architecture of ALAA, discusses the algorithm and controller design, and verifies the feasibility of the scheme and design through experiments with the physical prototype, showing its performance in different modes.

Robotic crop phenotyping has emerged as a key technology for assessing crops' phenotypic traits at scale, which is essential for developing new crop varieties with the aim of increasing productivity and adapting to the changing climate. However, developing and deploying crop phenotyping robots faces many challenges, such as complex and variable crop shapes that complicate robotic object detection, dynamic and unstructured environments that confound robotic control, and real-time computing and managing big data that challenge robotic hardware/software. This work specifically addresses the first challenge by proposing a novel Digital Twin(DT)/MARS-CycleGAN model for image augmentation to improve our Modular Agricultural Robotic System (MARS)'s crop object detection from complex and variable backgrounds. The core idea is that in addition to the cycle consistency losses in the CycleGAN model, we designed and enforced a new DT/MARS loss in the deep learning model to penalize the inconsistency between real crop images captured by MARS and synthesized images generated by DT/MARS-CycleGAN. Therefore, the synthesized crop images closely mimic real images in terms of realism, and they are employed to fine-tune object detectors such as YOLOv8. Extensive experiments demonstrate that the new DT/MARS-CycleGAN framework significantly boosts crop/row detection performance for MARS, contributing to the field of robotic crop phenotyping. We release our code and data to the research community (https://github.com/UGA-BSAIL/DT-MARS-CycleGAN).

A compliant plate driven by an active joint is frequently employed as a fin to improve swimming efficiency due to its continuous and compliant kinematics. However, very few studies have focused on the performance-oriented design of multijoint mechanisms enhanced with flexible fins, particularly regarding critical design factors such as the active-joint ratio and dimension-related stiffness distribution of the fin. To this aim, we developed a robotic tadpole by integrating a multijoint mechanism with a flexible fin and conduct a comprehensive investigation of its swimming performance with different tail configurations. A dynamic model with identified hydrodynamic parameters was established to predict propulsive performance. Numerous simulations and experiments were conducted to explore the impact of the active-joint ratio and the dimension-related stiffness distribution of the fin. The results reveal that (a) tails with different active-joint ratios achieve their best performance at a small phase difference, while tails with a larger active-joint ratio tend to perform worse than those with a smaller active-joint ratio when a larger phase difference is used; (b) the optimal active-joint ratio enables the robot to achieve superior performance in terms of swimming velocity and energy efficiency; and (c) with the same surface area, a longer fin with a wide leading edge and a narrow trailing edge can achieve higher swimming speeds with lower energy consumption. This work presents novel and in-depth insights into the design of bio-inspired underwater robots with compliant propulsion mechanisms.