Advanced intelligent systems (Weinheim an der Bergstrasse, Germany)最新文献

Fluid-driven artificial muscles exhibit a behavior similar to biological muscles which makes them attractive as soft actuators for wearable assistive robots. However, state-of-the-art fluidic systems typically face challenges to meet the multifaceted needs of soft wearable robots. First, soft robots are usually constrained to tethered pressure sources or bulky configurations based on flow control valves for delivery and control of high assistive forces. Second, although some soft robots exhibit untethered operation, they are significantly limited to low force capabilities. Herein, an electrohydraulic actuation system that enables both untethered and high-force soft wearable robots is presented. This solution is achieved through a twofold design approach. First, a simplified direct-drive actuation paradigm composed of motor, gear-pump, and hydraulic artificial muscle (HAM) is proposed, which allows for a compact and lightweight (1.6 kg) valveless design. Second, a fluidic engine composed of a high-torque motor with a custom-designed gear pump is created, which is capable of generating high pressure (up to 0.75 MPa) to drive the HAM in delivering high forces (580 N). Experimental results show that the developed fluidic engine significantly outperforms state-of-the-art systems in mechanical efficiency and suggest opportunities for effective deployment in soft wearable robots for human assistance.

Loading individual oocytes is a critical step for injecting RNA, expressing heterologous proteins, performing electrophysiological measurements, and so on. However, existing methods remain a challenge for automatically loading multiple single oocytes into different locations. Herein, a novel microfluidic chip on a robotic manipulator (chip-on-robot) with feedback control for flexible manipulation of multiple oocytes within a large spatial range is proposed. The manipulator automatically controls the microfluidic chip to reach different locations based on imaging feedback. The microfluidic chip then utilizes the hydrodynamic focusing effect of the main channel to separate oocytes for individual loading or reloading under capacitive sensor feedback. The separation distance reaches approximately 16 times the oocyte diameter. Moreover, capacitive signal feedback on the number of oocytes for flow direction control ensures the separation of all oocytes. For close-loop control of the loading/reloading process, image-based oocyte detection is combined using deep learning to calculate the target position of the oocyte. Finally, an automatic sequence is achieved to load multiple single oocytes into a well chip by using the chip-on-robot. As a demonstration, the oocytes are reloaded into a specified location based on the conditions. The proposed chip-on-robot with feedback control has significant advantages in the micromanipulation of oocytes.

Tactile sensing in the human body is achieved via the skin. This has inspired the fabrication of synthetic skins with pressure sensors for potential applications in robotics, bio-medicine, and human–machine interfaces. Tactile sensors based on magnetic elements are promising as they provide high sensitivity and a wide dynamic range. However, current magnetic tactile sensors mostly detect pressures of solid objects and operate at relatively high forces about 100 mN. Herein, these limitations are addressed by manufacturing soft, stretchable, and hair-like structures that are permanently magnetized to achieve high-resolution, cost-effective, and high-resolution pressure sensing. Combining these hair-like structures with advances in 3D magnetic-field measurements allows us to monitor directional tactile pressures without solid contact. To prove the concept of this technology, a bio-inspired soft device is built with a hairy structure that senses and reports environmental mechanical stresses, similar to that of human skin. Simple self-assembly of the soft magnetic hair structure makes our approach easy to scale for large-area applications.

Two of the main challenges in origami antenna designs are creating a reliable hinge and achieving precise actuation for optimal electromagnetic (EM) performance. Herein, a waterbomb origami ring antenna is introduced, integrating the waterbomb origami principle, 3D-printed liquid metal (LM) hinges, and robotic shape morphing. The approach, combining 3D printing, robotic actuation, and innovative antenna design, enables various origami folding patterns, enhancing both portability and EM performance. This antenna's functionality has been successfully demonstrated, displaying its communication capabilities with another antenna and its ability to navigate narrow spaces on a remote-controlled wheel robot. The 3D-printed LM hinge exhibits low DC resistance (200 ± 1.6 mΩ) at both flat and folded state, and, with robotic control, the antenna achieves less than 1° folding angle accuracy and a 66% folding area ratio. The antenna operates in two modes at 2.08 and 2.4 GHz, ideal for fixed mobile use and radiolocation. Through extensive simulations and experiments, the antenna is evaluated in both flat and folded states, focusing on resonant frequency, gain patterns, and hinge connectivity. The findings confirm that the waterbomb origami ring antenna consistently maintains EM performance during folding and unfolding, with stable resonant frequencies and gain patterns, proving the antenna's reliability and adaptability for use in portable and mobile devices.

Although large-scale pretrained convolutinal neural networks (CNN) models have shown impressive transfer learning capabilities, they come with drawbacks such as high energy consumption and computational cost due to their potential redundant parameters. This study presents an innovative weight-level pruning technique that mitigates the challenges of overparameterization, and subsequently minimizes the electricity usage of such large deep learning models. The method focuses on removing redundant parameters while upholding model accuracy. This methodology is applied to classify Eimeria species parasites from fowls and rabbits. By leveraging a set of 27 pretrained CNN models with a number of parameters between 3.0M and 118.5M, the framework has identified a 4.8M-parameter model with the highest accuracy for both animals. The model is then subjected to a systematic pruning process, resulting in an 8% reduction in parameters and a 421M reduction in floating point operations while maintaining the same classification accuracy for both fowls and rabbits. Furthermore, unlike the existing literature where two separate models are created for rabbits and fowls, this article presents a combined model with 17 classes. This approach has resulted in a CNN model with nearly 50% reduced parameter size while retaining the same accuracy of over 90%.

Insects flap their wings through a highly specialized musculoskeletal system that allows the wings to rotate about three degrees of freedom. Consequently, the wingtip trajectory is adjustable in 3D, and accompanied with appropriate wing feathering (wing pitch). Remarkably, the complex flapping motion is achieved by thoracic muscles acting on the wing hinge. The wings themselves do not possess muscles but adjust their shape and orientation by elastically deforming due to the loads applied on them during flapping. Previous attempts to develop insect-inspired flapping drones have mostly focused on simplified linear flapping mechanisms, which do not utilize the interaction between the wing flexibility and flapping kinematics to its full potential. Here, the aim is to improve flapping drones’ performance by introducing mechanisms that mimic insects’ flight. The first is an elastic beam mechanism, allowing the wing root to swing during flapping, and the second is a passive wing pitch mechanism that allows the wing to rotate at stroke reversals. The two mechanisms are tested using high-fidelity insect-inspired 3D-printed wings and show a sixfold improvement of aerodynamic performance compared to linear flapping kinetics of the same flexible wings. This underscores the necessity of bioinspired flapping mechanisms in future flapping drones.

Due to the absence of more efficient diagnostic tools, the spread of mpox continues to be unchecked. Although related studies have demonstrated the high efficiency of deep learning models in diagnosing mpox, key aspects such as model inference speed and parameter size have always been overlooked. Herein, an ultrafast and ultralight network named Fast-MpoxNet is proposed. Fast-MpoxNet, with only 0.27 m parameters, can process input images at 68 frames per second (FPS) on the CPU. To detect subtle image differences and optimize model parameters better, Fast-MpoxNet incorporates an attention-based feature fusion module and a multiple auxiliary losses enhancement strategy. Experimental results indicate that Fast-MpoxNet, utilizing transfer learning and data augmentation, produces 98.40% classification accuracy for four classes on the mpox dataset. Furthermore, its Recall for early-stage mpox is 93.65%. Most importantly, an application system named Mpox-AISM V2 is developed, suitable for both personal computers and smartphones. Mpox-AISM V2 can rapidly and accurately diagnose mpox and can be easily deployed in various scenarios to offer the public real-time mpox diagnosis services. This work has the potential to mitigate future mpox outbreaks and pave the way for developing real-time diagnostic tools in the healthcare field.

Bionic microrobots working in liquid environments have attracted attention in recent years, because they play an important role in the medical fields. So far, most bionic microrobots serving in liquid environments (swimming microrobots) are fabricated based on organic materials. Limited by the inherent property of organic materials, the performance and lifetime of the swimming microrobots are still deficient. Facing this challenge, inspired by sperms, swimming microrobots based on the inorganic phase transition driving material vanadium dioxide are developed. In liquid environments, the linear and rotary motion of these sperm-like micro-robots could be controlled by changing the laser modulation frequency. The highest linear speed attained is 56 μm s⁻¹, and the highest rotary speed attained is 14° s⁻¹. The microrobot is able to undergo more than 10⁵ cycles in a liquid environment without degradation of its performance. Considering its high performance and controllability, the swimming microrobot is expected to be helpful in medical applications such as precision drug delivery and minimally invasive surgery.

Soft and lightweight grippers have greatly enhanced the performance of robotic manipulators in handling complex objects with varying shape, texture, and stiffness. However, the combination of universal grasping with passive sensing capabilities still presents challenges. To overcome this limitation, a fluidic soft gripper is introduced based on the buckling of soft, thin hemispherical shells. Leveraging a single fluidic pressure input, the soft gripper can grasp slippery and delicate objects while passively providing information on this physical interaction. Guided by analytical, numerical, and experimental tools, the novel grasping principle of this mechanics-based soft gripper is explored. First, the buckling behavior of a free hemisphere is characterized as a function of its geometric parameters. Inspired by the free hemisphere's two-lobe mode shape ideal for grasping purposes, it is demonstrated that the gripper can perform dexterous manipulation and gentle gripping of fragile objects in confined spaces and underwater environments. Last, the soft gripper's embedded capability of detecting contact, grasping, and release conditions during the interaction with an unknown object is proved. This simple buckling-based soft gripper opens new avenues for the design of adaptive gripper morphologies with tactile sensing capabilities for applications ranging from medical and agricultural robotics to space and underwater exploration.

The obscure drug response continues to be a limiting factor for accurate cures for cancer. Next generation sequencing technologies have propelled the pharmacogenomic studies with characterized large panels of cancer cell line at multi-omics level. However, the sufficient integration of the multi-omics data and the efficient prediction for drug response and synergy still remain a challenge. To address these problems, ECFD is designed, an ensemble cascade forest-based framework that predicts drug response and synergy using five types of omics data. Experimental results show the significant advantages of the ECFD model over existing models. The best integration of feature extraction is determined and the superiorities of robust stability in the face of new and small samples are highlighted. In addition, the methodological framework highlights the explainability of the model, the mechanisms of drug resistance and drug combination treatment strategies based on explainable analyses and biological networks. In sum, ECFD may facilitate the evaluation of drug response and speculation of potential synergy therapies in personalized and precision treatment.