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

Event cameras are novel bioinspired vision sensors that measure pixel-wise brightness changes asynchronously instead of images at a given frame rate. They offer promising advantages, namely, a high dynamic range, low latency, and minimal motion blur. Modern computer vision algorithms often rely on artificial neural network approaches, which require image-like representations of the data and cannot fully exploit the characteristics of event data. Herein, approaches to action recognition based on the Fourier transform are proposed. The approaches are intended to recognize oscillating motion patterns commonly present in nature. In particular, the approaches are applied to a recent dataset of breeding penguins annotated for “ecstatic display,” a behavior where the observed penguins flap their wings at a certain frequency. It is found that the approaches are both simple and effective, producing slightly lower results than a deep neural network (DNN) while relying just on a tiny fraction of the parameters compared to the DNN (five orders of magnitude fewer parameters). They work well despite the uncontrolled, diverse data present in the dataset. It is hoped that this work opens a new perspective on event-based processing and action recognition.

Soft manipulators, renowned for their compliance and adaptability, hold great promise in their ability to engage safely and effectively with intricate environments and delicate objects. Nonetheless, controlling these soft systems presents distinctive hurdles owing to their nonlinear behavior and complicated dynamics. Learning-based controllers for continuum soft manipulators offer a viable alternative to model-based approaches that may struggle to account for uncertainties and variability in soft materials, limiting their effectiveness in real-world scenarios. Learning-based controllers can be trained through experience, exploiting various forward models that differ in physical assumptions, accuracy, and computational cost. In this article, the key features of popular forward models, including geometrical, pseudo-rigid, continuum mechanical, or learned, are first summarized. Then, a unique characterization of learning-based policies, emphasizing the impact of forward models on the control problem and how the state of the art evolves, is offered. This leads to the presented perspectives outlining current challenges and future research trends for machine-learning applications within soft robotics.

Fluidic control systems target unique applications where conventional electronics fail. However, current fluidic control systems face challenges in accessible fabrication, reproducibility, and modifiable characteristics such as operating pressure and instability count. Herein, fused deposition-modeled compliant mechanisms with flexing beams and soft linear actuators for fluid switching and the control of soft robotic systems are introduced. A linear actuator switches a compliant mechanism to cut off airflow through off-the-shelf tubing. The modular compliant logic devices can be configured as normally open or normally closed switches, as NOT, AND, and OR gates, and as nonvolatile memory elements. Their use is demonstrated in controlling a fluidic stepper motor, a worm-like robot, and a fluidic display. These fluidic switches are printable using inexpensive desktop 3D printers, can be reliably reproduced in large quantities, and offer a wide range of modifiable parameters, including scalability, adaptability in operating pressure, and the tunability of instability counts for computational and memory functions.

Direct printing methods are widely recognized as efficient techniques for manufacturing printed electronics. However, several challenges arise when printing on nonplanar surfaces, especially using the drop-on-demand (DoD) approach. These challenges include ink flow due to gravity, precise ink deposition, and reproducibility. This study introduces an innovative method for highly accurate DoD material jetting on nonplanar 3D conductive surfaces, enabling precise production and trajectory control of charged droplets. The technique involves using a grounded 3D substrate as the target, where in-flight droplets are subjected to an external electric field generated by gate electrode installed on a piezo activated droplet dispenser. Individual droplets are generated and controlled using a complex trigger system that relays variable-voltage signals to the gate electrode. Moreover, a predictive model for droplet deposition, exhibiting an accuracy of 87%, is developed utilizing supervised machine learning (ML). This approach significantly improves the accuracy and repeatability of droplet deposition. Overall, this study presents an effective method of integrating piezoelectric and electrohydrodynamic printing technologies, complemented by ML. It addresses the challenges associated with printing on nonplanar surfaces using the DoD material jetting technique and shows considerable promise for enhancing efficiency, accuracy, and repeatability in the manufacturing of printed electronics.

Sorptive bioprocesses are the basis for numerous biotechnological applications such as enzyme immobilization, biosensors, controlled drug delivery, water treatment, and molecular purification. Yet due to the complexity of these processes, their optimization is still time, labor, and cost-intensive. This research presents a flexible self-driving laboratory (SDL) designed for the accelerated development and optimization of solid-phase extraction processes. As a use case, the SDL was used to optimize a DNA purification process using silica magnetic beads. Through the integration of robotics, machine learning, and data-driven experimentation, the SDL demonstrates a highly accelerated process optimization with minimal human intervention. In the multistep purification approach, the system is able to optimize buffer compositions for DNA extraction from complex samples, demonstrating effectiveness in both conventional chaotropic salt-based methods and innovative chaotropic salt-free buffers. The study highlights the SDL's capability to autonomously refine process parameters, achieving significant enhancements in yield and purity of the product. This blueprint for future self-driving optimization of bioprocess parameters showcases the potential of autonomous systems to revolutionize biochemical process development, offering insights into scalable, environmentally sustainable, and cost-effective solutions.

The robotic manipulation of deformable linear objects is a frontier problem with many potential applications in diverse industries. However, most existing research in this area focuses on shape control for a provided explicit goal and does not consider physical constraints, which limits its applicability in many real-world scenarios. In this study, a self-supervised planning and control approach are proposed to address the challenge of rearranging deformable linear objects for implicit goals. Specifically, the context of making both ends of the object reachable (inside the robotic access range) and graspable (outside potential collision regions) by dual-arm robots is considered. Firstly, the object is described with sequential keypoints and the correspondence-based action is parameterized. Secondly, a generator capable of producing multiple explicit targets is developed, which adhere to implicit conditions. Thirdly, value models are learnt to assign the most promising explicit target as guidance and determine the goal-conditioned action. All models within the policy are trained in a self-supervised manner based on data collected from simulations. Importantly, the learned policy can be directly applied to real-world settings since we do not rely on accurate dynamic models. The performance of the new method is validated with simulations and real-world experiments.

First responders often face hazardous and life-threatening situations in environments filled with smoke, posing significant risks to their safety. The existing perception solutions, such as camera or light detection and ranging (LiDAR)-based methods, are inadequate when faced with visually degraded conditions caused by smoke. In this work, SmokeNav, a novel system that combines data from an inertial sensor and millimeter-wave (mmWave) radar, is proposed to enhance situational awareness for first responders in smoky environments. SmokeNav utilizes an inertial positioning module that exploits the human motion constraints with a foot-mounted inertial measurement unit to provide accurate user localization. By integrating this location information with mmWave radar data, it employs a probabilistic occupancy map construction to reconstruct an accurate metric map. To enable semantic understanding of the environment, a DNN-based semantic segmentation model that incorporates radar reflectivity and employs focal loss to improve performance is introduced. Herein, extensive real-world experiments in smoky environments is conducted to demonstrate that SmokeNav precisely localizes the user and generates detailed maps with semantic segmentation. In this work, potentials are held for enhancing the safety and effectiveness of first responders in hazardous conditions.

Robotic grasping and manipulation demand the ability to handle a multitude of objects with different shapes, sizes, quantities, surface smoothness, vulnerability, and stiffness, which is challenging without prior knowledge about object properties. Herein, a novel origami-inspired gripper for universal grasping is presented. The innovative structure seamlessly transforms a simple uniaxial pulling motion into a flexible and robust envelope or pinch grasp, enabling it to tackle various scenarios. The origami gripper offers distinctive advantages, including scalable and optimizable design, grasping compliance and robustness, providing material flexibility, and providing solutions to challenging manipulation tasks. The working principles of the origami gripper are characterized and analyzed. An optimization-based inverse design method is presented to adjust gripper properties for various scenarios. Through comprehensive experimentation and evaluation, the gripper's capabilities to grasp various objects with a wide range of distinctive properties, including ultrasoft, slippery, granular, and multiple objects, which is a challenge for the existing robotic grippers, are demonstrated. The research holds promise for transformative applications in areas such as the food industry, waste handling, fine and fragile objects grasping, and environmental sampling.

The star-nosed mole is recognized as a tactilely sensitive mammal due to its unique nose, which facilitates spatial detection in dark environments through touch using its appendages enveloped in numerous sensory receptors. This article introduces a bionic soft robot inspired by the star-nosed mole, which combines a pneumatic soft platform with a polydimethylsiloxane–polyethylene terephthalate cylindrical tactile sensor array based on bilayer single-electrode triboelectric nanogenerators, mimicking the muscle tissue of the mole's nose and the cylindrical appendages surrounded by Eimer's organs. The cylindrical sensor array enables multiangle spatial detection without an external power supply and remains unaffected by external materials. By implementing a constant curvature model for robot motion control, positional information is provided for the contact points between the cylindrical sensor array and the external environment. The robot effectively discriminates the distance and shape of various objects and achieves nonvisual 3D spatial detection and reconstruction in real-world scenarios. This work presents a novel bionic approach for 3D spatial detection in nonvisual environments.

Transrectal ultrasound (TRUS) videos offer valuable histopathologic information about the prostate. Accurate prostate zonal segmentation in TRUS videos is vital for diagnosing prostate cancer and guiding surgery. However, TRUS videos are manually recorded by urologists, resulting in no standardized coordinate system, which limits direct prostate zonal segmentation in these videos. To overcome the limitation, a novel Prostate Zonal Segmentation Network (PZS-Net), based on U-Net, which learns critical cross-frame information and multi-scale features from sequential frames, is proposed. First, a sequential frame cross-attention (SFCA) module is designed to capture remote information from sequential frames to enhance the feature representation of the current frame. The SFCA module is embedded at each skip connection layer to extract crucial cross-frame information. Then, a multi-scale fusion (MSF) module that utilizes three parallel branches with different atrous convolutions is designed. The MSF module is placed at the bottleneck layer to dynamically fuse multi-scale context information from high-level features. Extensive experiments on TRUS image datasets show that the PZS-Net achieves higher accuracy in both the transitional zone (dice coefficient [Dice]: 68.90% ± 1.73%, mean intersection over union [mIoU]: 59.19% ± 2.09%, 95% Hausdorff distance [HD95]: 5.02 ± 0.83 mm) and the peripheral zone (Dice: 63.99% ± 3.16%, mIoU: 54.60% ± 3.35%, HD95: 5.28 ± 1.12 mm) and demonstrates the effectiveness and competitiveness of its key components via comprehensive ablation studies.