High-level Input Research Articles

From Surgical Procedures to Glovebox Processing: Enhancing Safety and Efficiency in Glovebox Operations for High-Consequence Materials

High-consequence materials, such as nuclear waste, pose significant risks if improperly handled. Gloveboxes provide a controlled environment to manage these materials safely, yet they present challenges in operator safety, efficiency, and ergonomics. This paper addresses these challenges by presenting a proof-of-concept framework for Semi-Autonomous Robot Teleportation for Nuclear Glovebox Processing (SART-NGP). Unlike the existing manual systems already in use at nuclear facilities, the SART-NGP framework incorporates adjustable autonomy capabilities, drawing inspiration from advancements in surgical robotics. By integrating high-level human inputs, semi-autonomous execution, robust feedback mechanisms, and a simulator-based interface, this framework significantly enhances safety and efficiency while alleviating ergonomic concerns associated with glovebox processing in nuclear facilities. Although we have not fully solved the challenges associated with nuclear glovebox processing, our research provides promising insights and methodologies that may significantly improve the management of high-consequence materials and augments existing efforts in self-driving labs for glovebox operations. The findings suggest that the SART-NGP framework holds substantial potential for enhancing glovebox processing in nuclear facilities, with future work focusing on its real-world implementation and evaluation.

The Problem of Sequential Entry of Nurses as a Disincentive to Medical Safety and a Factor Increasing Overtime.

The purpose of this study was to visualize the nurses' approach to recording and to identify the causes that lead to overtime. The data used were those related to the performance and entry of nursing tasks (observation and care acts). Data from Hospital A, a provider of acute care with approximately 400 beds, was used. One month of work data were obtained from 12 nurses who were judged by the chief nurse to be excellent and efficient in their work. 12 nurses were divided into three groups: a high-level sequential entry rate group (>0.6), a medium-level sequential entry rate group (0.4 to <0.6), and a low-level sequential entry rate group (<0.4). The high-level sequential input rate group had data input every time they performed their work, so that the work implementation and input implementation were completed at 5:00 p.m. The medium-level sequential entry rate group tended to enter data about an hour late when they started work, which tended to result in overtime. The low-level sequential input rate group tended to input work in the afternoon, resulting in significant overtime. The data of approximately 400 nurses in a 900-bed University Hospital B, where Team Compass is implemented, were used to analyze the actual sequential input rate. 884 nurses in University Hospital B, of whom 397 had input data for 19 consecutive months. The sequential input rate in November 2020, one and a half years after the start of operation, was divided into five groups by 20%: 0-20%: 27 (7%), 20-40%: 30 (7%), 40-60%: 37 (9%), 60-80%: 69 (17%), and over 80%: 234 (59%). The number of employees who have been working overtime hours is also high. Although the importance of sequential entry to achieve results in reducing overtime hours continued to be strongly presented, there were 23% of nurses who indicated a sequential entry rate of less than 60%.

Reliability Assessment of 60-GHz GaN Power Amplifier Under High-Level Input RF Stress

Reliability Assessment of 60-GHz GaN Power Amplifier Under High-Level Input RF Stress

Flooding soil with biogas slurry suppresses root-knot nematodes and alters soil nematode communities

Root-knot nematodes (RKNs) pose a serious threat to crop production. Flooding soil with biogas slurry, combined with soil heating before crop planting, has the potential for RKN disease suppression. However, the actual effect of this method has not been verified under field conditions. Here, we present the results of a two-year field experiment in a greenhouse demonstrating the control effect on RKN disease and plant growth using this method, as well as its influence on the soil nematode community. Four treatments were set: untreated control (CK), local control method for RKN (CC), soil flooded with 70 % biogas slurry (BS70), and soil flooded with undiluted biogas slurry (BS100). In the first year, all three RKN control treatments significantly reduced the root-knot index (p < 0.05). In the next year, only BS70 and BS100 still presented significantly suppressed effects (p < 0.05), and it was more obvious under BS70 with a relative control effect of 74.6 %. In the first year, BS70 and BS100 significantly inhibited the plant height of watermelon (p < 0.05). In the next year, however, all three RKN control treatments promoted the growth of watermelon, and their stem diameter was significantly greater than that of CK. The application of biogas slurry (BS70 and BS100) significantly increased nematode richness and the Shannon index in the second year (p < 0.05). However, the structure index showed no significant difference among treatments (p > 0.05), indicating that biogas slurry application did not increase the soil food web complex. Principal component analysis showed that the application of biogas slurry changed the nematode community, especially under BS70, which presented a more lasting influence. The high-level input of biogas slurry also caused soil NH4+-N and heavy-metal and arsenic accumulation in the first year, but these soil-pollution risks disappeared in the second year.

Improving new physics searches with diffusion models for event observables and jet constituents

We introduce a new technique called Drapes to enhance the sensitivity in searches for new physics at the LHC. By training diffusion models on side-band data, we show how background templates for the signal region can be generated either directly from noise, or by partially applying the diffusion process to existing data. In the partial diffusion case, data can be drawn from side-band regions, with the inverse diffusion performed for new target conditional values, or from the signal region, preserving the distribution over the conditional property that defines the signal region. We apply this technique to the hunt for resonances using the LHCO di-jet dataset, and achieve state-of-the-art performance for background template generation using high level input features. We also show how Drapes can be applied to low level inputs with jet constituents, reducing the model dependence on the choice of input observables. Using jet constituents we can further improve sensitivity to the signal process, but observe a loss in performance where the signal significance before applying any selection is below 4σ.

Comparing Direct-to-Consumer Devices to Hearing Aids: Amplification Accuracy for Three Types of Hearing Loss.

The U.S. Food and Drug Administration finalized regulations for over-the-counter hearing aids (OTC-HAs) on August 17, 2022. Little is known about the comparative performance of OTC-HAs and prescription HAs. This study compared amplification accuracy of prescription HAs and direct-to-consumer devices (DTCDs, including personal sound amplification products [PSAPs] and OTC-HAs). Eleven devices were programmed to meet prescriptive targets in an acoustic manikin for three degrees of hearing loss. Devices consisted of high- and low-end HAs, PSAPS, and OTC-HAs. Each was tested, and deviations from target measured with an HA analyzer at every combination of 10 frequencies and low-, average-, and high-level inputs. Accuracy was compared using a multilevel Poisson model with device-specific intercepts controlling for input level, frequency, and device type. For mild-moderate hearing loss, deviations from targets were not statistically different between high- and low-end HAs, but PSAPs (5.50 dB, SE = 0.92 dB) and OTC-HAs (8.83 dB, SE = 1.10 dB) had larger differentials. For flat moderate hearing loss, compared to high-end HAs, average differentials were larger for all device types at all input levels and frequencies (Low HA: 3.82 dB, SE = 1.10 dB; PSAP: 9.24 dB, SE = 1.22 dB; OTC-HA: 8.61 dB, SE = 1.19 dB). For mild sloping to severe hearing loss, compared to high-end HAs, OTC-HAs (9.72 dB, SE = 1.20 dB) and PSAPs (7.34 dB, SE = 1.07 dB) had larger differentials and significant variability at the highest and lowest frequencies. Half (three) of the PSAPs and OTC-HAs met most targets within ±5 dB. DTCDs were unable to meet prescriptive targets for severe types of hearing loss but could meet them for mild hearing loss. This study provides an examination of current hearing devices. More research is needed to determine whether meeting prescriptive targets provides any benefit in the outcomes and performance with DTCD devices.

Trajectory Tracking Outer Loop Regressor-Based Adaptive Controller for a Quadrotor

A high-level control strategy for a quad rotorcraft Unmanned Aircraft System to perform trajectory tracking tasks is presented, which is based on a regressor-based adaptive approach. The high-level control is designed to interact with a low-level (internal) control loop that cannot be modified to suit the needs of academic researchers. Hence, the proposed control framework computes the appropriate high-level inputs for the inner controller, enabling the trajectory tracking task. The controller includes an integral action to overcome steady-state errors that may occur due to parameter estimation errors or constant disturbances. The stability of the equilibrium point is analyzed using Lyapunov theory, which shows that the tracking errors converge to zero and the parameter estimation errors remain bounded. The proposed control framework was tested on a real-time quad rotorcraft platform, and its performance was compared with four different control strategies. The results indicate that the proposed controller exhibits high accuracy and has better performance with respect to the other controllers.

Language support for verifying reconfigurable interacting systems

Reconfigurable interacting systems consist of a set of autonomous agents, with integrated interaction capabilities that feature opportunistic interaction. Agents seemingly reconfigure their interaction interfaces by forming collectives and interact based on mutual interests. Finding ways to design and analyse the behaviour of these systems is a vigorously pursued research goal. In this article, we provide a modelling and analysis environment for the design of such system. Our tool offers simulation and verification to facilitate native reasoning about the domain concepts of such systems. We present our tool named R-CHECK (please find the associated toolkit repository here: https://github.com/dsynma/recipe). R-CHECK supports a high-level input language with matching enumerative and symbolic semantics and provides modelling convenience for features such as reconfiguration, coalition formation, and self-organisation. For analysis, users can simulate the designed system and explore arising traces. Our included model checker permits reasoning about interaction protocols and joint missions.

Digital Economy and Intelligent Manufacturing Coupling Coordination: Evidence from China

The digital economy uses its own digital information advantages to reduce the intensity of energy consumption brought by economic growth. Intelligent manufacturing achieves cost reduction and efficiency through the integration of manufacturing and intelligence as well as digitalization and information technology. The two have become a new engine for sustainable economic development at present, and they can promote and influence each other. However, there is a lack of research on the relationship between them. In this regard, this study aims to build a coupling coordination model of digital economy and intelligent manufacturing and to make an empirical analysis using the data of Chinese provincial administrative regions in order to provide a theoretical reference for promoting sustainable economic development. The research finds that (1) the digital economy and intelligent manufacturing are mainly cross-coupled from four aspects: infrastructure, technological innovation, product optimization and organizational change. The development level and speed of the former are significantly higher than those of the latter, and the gap does not decrease with time. The two have a strong correlation, but there is no high-quality coupling coordination. (2) The main obstacle factors to the digital economy lie in the imperfect supporting facilities, the short board of technological innovation and the lack of technological application capacity. Intelligent manufacturing lacks intelligent application and technological innovation. (3) Influencing factors such as opening to the outside world, economic development, high-level talent input, industrial structure and innovation emphasis have different effects on their coupling and coordinated development in different regions. (4) The spatial correlation test shows that the coupling coordination degree of each region is spatially positively correlated. This research helps to promote the coupling and coordinated development of the digital economy and intelligent manufacturing.

Generative modelling of vibration signals in machine maintenance

The exponential development of technologies for the acquisition, collection, and processing of data from real-world objects is creating new perspectives in the field of machine maintenance. The Industrial Internet of Things is the source of a huge collection of measurement data. The performance of classification or regression algorithms needs to take into account the random nature of the process being modelled and any incomplete observability, especially in terms of failure states. The article highlights the practical possibilities of using generative artificial intelligence and deep machine learning systems to create synthetic measurement observations in monitoring the vibrations of rotating machinery to improve unbalanced databases. Variational Autoencoder VAE generative models with latent variables in the form of high-level input features of time-frequency spectra were studied. The mapping and generation algorithm was optimised and its effectiveness was tested in the practical solution of the task of diagnosing the three operating states of a demonstration gearbox.

AN OVERVIEW ON SUGARCANE INDUSTRY IN INDIA

Sugarcane is an important industrial and cash crop. It is grown in almost in all areas in India, But commercial basis sugarcane is produced in tropical and sub tropical states. Because of available of raw materials, most of the sugar mills are located in these areas. Sugarcane production in India has been attributes to lack of rainfall, reduction in cultivation area. Also production of sugarcane affects due to drought and decrease in soil health, high level input cost. This paper presents an overview of sugar industries, state wise area of cultivation, production, productivity. This paper is descriptive in nature and completely based on the data and information available from secondary sources. Time frame of the study is ten years from 1st April, 2012 to 31st March, 2022. KEYWORDS – Production of Sugarcane, sugarcane cultivation, export &amp; import of Sugar, fair and remunerative price of Sugar

Missing Child Identification System using Deep Learning and Multiclass SVM for E-police

Countless children go missing in India every year. In missing children, most children are not included. This article presents a deep learning method to identify missing children from multiple existing images of children with the help of face recognition. Citizens can send pictures of suspicious children to public places with signs and writings. The photo will be automatically compared to the saved photo of the missing child in the warehouse. Sort the photos of the children and choose the best photos from the missing children file. To this end, a deep learning model was trained to accurately identify missing children by matching information about missing children using face images provided by the public. Convolutional Neural Networks (CNNs) are deep learning techniques for image-based applications, this is face recognition. Extract face identifiers from images using the CNN pre-trained VGG-Face deep architecture. Compared to ordinary deep learning applications, our algorithm only uses network connection as high-level input and child identification is done by SVM trainers. Choose VGG-Face, the most effective CNN face recognition model, and train it appropriately to make deep learning models that are not affected by noise, light, contrast, obscuration, image posture and child age, and perform better based on previous Face recognition methods. In identifying missing children, Child identification achieved a 99.41% classification rate. Forty-three patients were evaluated

Fast Obstacle Avoidance Based on Real-Time Sensing

Humans excel at navigating and moving through dynamic and complex spaces, such as crowded streets. For robots to do the same, it is crucial that they are endowed with highly reactive obstacle avoidance which is adept at partial and poor sensing. We address the issue of enabling obstacle avoidance based on sparse and asynchronous perception. The proposed control scheme combines a high-level input command provided by either a planner or a human operator with fast reactive obstacle avoidance (FOA). The sampling-based sensor data can be combined with an analytical reconstruction of the obstacles for real-time collision avoidance. Thus, we can ensure that the agent does not become stuck when a feasible path exists between obstacles. Our algorithm was evaluated experimentally on static laser data from cluttered, indoor office environments. Additionally, it was used in shared-control mode in a dynamic and complex outdoor environment in the center of Lausanne. The proposed control scheme successfully avoided collisions in both scenarios. During the experiments, the controller took 1 ms to evaluate over 30000 data points.

A Constrained Control Allocation and Tuning Scheme for Hybrid Actuators in Spacecraft Attitude Control

Agile rotational maneuvers of spacecraft requires careful execution since its actuators may not be able to produce the demanded torques, causing the state trajectories to deviate and a desired attitude would not be guaranteed. We investigate the control allocation problem for a redundant set of hybrid actuators that include reaction wheels, magnetorquers, and continuous-force thrusters. The main objective of the magnetorquers is to dump momentum from the reaction wheels, whereas the wheels are the primary actuators in attitude control, and more agile maneuvers or faster unloading of momentum can be handled by the thrusters. A modified mixed optimization scheme for control allocation is presented where the equality constraints account for satisfying the high-level (virtual) control inputs for both attitude control and momentum dumping. Variants of dynamic weights in the optimization are developed such that magnetorquers and thrusters may contribute with a relative degree of importance in the attitude control problem. The control allocation scheme is solved using quadratic programming where simulation results are shown for fast and slow rotational maneuvers together with fast and slow momentum dumping.

Measuring nonlinearities of a cantilever beam using a low-cost efficient wireless intelligent sensor for strain (LEWIS-S)

In the context of experimental vibration data, strain gauges can obtain linear and nonlinear dynamic measurements. However, measuring strain can be disincentivizing and expensive due to the complexity of data acquisition systems, lack of portability, and high costs. This research introduces the use of a low-cost efficient wireless intelligent sensor for strain (LEWIS-S) that is based on a portable-sensor-design platform that streamlines strain sensing. The softening behavior of a cantilever beam with geometric and inertial nonlinearities is characterized by the LEWIS-S based on high force level inputs. Two experiments were performed on a nonlinear cantilever beam with measurements obtained by the LEWIS-S sensor and an accelerometer. First, a sine sweep test was performed through the fundamental resonance of the system, then a ring-down test was performed from a large initial static deformation. Good agreement was revealed in quantities of interest such as frequency response functions, the continuous wavelet transforms, and softening behavior in the backbone curves.

Fabricatable axis: an approach for modelling customized fabrication machines

Digital fabrication tools such as 3D printers, computer-numerically controlled (CNC) milling machines, and laser cutters are becoming increasingly available, ranging from consumer to industrial versions. Recent studies have shown that users, ranging from researchers, to industry professionals, to hobbyists, are interested in modifying and changing the inherit workflows these tools provide. As an answer to this, these users are increasingly modifying and customizing their machines by changing the work envelope, adding different end-effectors, and creating their own fabrication workflows in software. However, customizing, modifying and creating digital fabrication machines and the workflows they provide require extensive knowledge within multiple different engineering domains and is non-trivial. In this article we present a model-driven approach that enables users to expand their digital fabrication scope by providing a high-level tool that facilitates the customization of fabrication tools. We present The Farbicatable Axis, a model that enables users to create customized linear actuators. The model takes high-level input parameters such as length and gearing-parameters, and outputs a CAD model of a linear motion axis consisting of fabricatable parts. We then present how instances of the Fabricatable Axis can be combined and used to design and implement Fabricatable Machines.

Effects of Green Manure Combined with Phosphate Fertilizer on Movement of Soil Organic Carbon Fractions in Tropical Sown Pasture

The application of green manure is a common way to increase the soil’s level of total organic carbon (TOC) and its fractions. However, the amount of green manure to apply and how the combined application of phosphate fertilizer affects the movement of TOC, and of its fractions, is still unclear. We conducted a column experiment with two treatments of phosphate fertilizer (with and without) and green manure (0, low amount level, high amount level). The longitudinal movement distance and accumulation amount of each organic carbon fraction were investigated after 14 days and 28 days. The results indicated that green manure, phosphate fertilizer, and incubation time affected the movement of the soil organic carbon fractions by affecting the initial quality of the green manure (TOC, cellulose, and lignin content), as well as the changes in quality. Green manure significantly increased the accumulation amount of the organic carbon fractions in the soil, and the high-level input of green manure increased the movement distance and accumulation amount of the organic carbon fractions; phosphorus fertilizer did not have a significant effect on the movement distance of the organic carbon fractions, but it did significantly affect accumulation amount. The 28-day incubation period increased the movement distance and the accumulation amount of the organic carbon fractions, with the exception of the particulate organic carbon (POC), compared to the 14-day incubation period. Taken together, these findings suggest that the high-level of input of green manure combined with the application of P fertilizer is beneficial for increasing the movement of the organic carbon fractions to the depth of the soil, and promotes their accumulation, which is an important agronomic management strategy for improving soil acidity in tropical regions.

Unsupervised monocular visual odometry via combining instance and RGB information.

Unsupervised deep learning methods have made significant progress in monocular visual odometry (VO) tasks. However, due to the complexity of the real-world scene, learning the camera ego-motion from the RGB information of monocular images in an unsupervised way is still challenging. Existing methods mainly learn motion from the original RGB information, lacking higher-level input from scene understanding. Hence, this paper proposes an unsupervised monocular VO framework that combines the instance and RGB information, named combined information based (CI-VO). The proposed method includes two stages. First is obtaining the instance maps of the monocular images, without finetuning on the VO dataset. Then we obtain the combined information from the two types of information, which is input into the proposed combined information based pose estimation network, named CI-PoseNet, to estimate the relative pose of the camera. To make better use of the two types of information, we propose a fusion feature extraction network to extract the fused features from the combined information. Experiments on the KITTI odometry and KITTI raw dataset show that the proposed method has good performance in the camera pose estimation task, which exceeds the existing mainstream methods.

Human-Scale Mobile Manipulation Using RoMan

We present the design, integration, and evaluation of a full-stack robotic system called RoMan, which can conduct autonomous field operations involving physical interaction with its environment. RoMan offers autonomous behaviors that can be triggered from succinct, high-level human input such as “open this box and retrieve the bag inside.” The robot’s behaviors are driven by a set of planners and controllers grounded in perceptual reconstructions of the environment. These behaviors are articulated by a behavior tree that translates high-level operator input into programs of increasing sensorimotor expressiveness, ultimately driving the lowest-level controllers. The software system is implemented in ROS as a set of independent processes connected by synchronous and asynchronous communication, and distributed across two on-board planning/control computers. The behavior stack drives a novel platform consisting of a pair of custom, 500 Nm/axis manipulators mounted on a rotatable torso aboard a tracked platform. The robot’s head is equipped with forward-looking depth cameras, and the arms carry wrist-mounted force-torque sensors and a mix of three- and four-finger grippers. We discuss design and implementation trade-offs affecting the entire hardware-software stack and high-level manipulation behaviors. We also demonstrate the applicability of the system for solving two manipulation tasks: 1) removing heavy debris from a roadway, where 64% of end-to-end autonomous runs required at most one human intervention; and 2) retrieving an item from a closed container, with a fully autonomous success rate of 56%. Finally, we indicate lessons learned and suggest outstanding research problems.

STL-Detector: Detecting City-Wide Ride-Sharing Cars via Self-Taught Learning

Ride-sharing cars are private vehicles held by individuals or provided by ride-hailing companies for designated drivers to offer taxi-like services. Recently, various ride-sharing cars have emerged around the city with the popularity of online ride-hailing services. Identifying them is the critical task of transportation management. However, less work focuses on this issue due to the lack of city-wide private vehicles&#x2019; trajectory data and labeled ride-sharing cars. Fortunately, data collected by advanced sensing technology, such as electronic registration identification (ERI) of the motor vehicle data collected by radio-frequency identification (RFID) technology, provide us with an opportunity to detect ride-sharing cars from a data-driven aspect. This article proposes detecting ride-sharing cars via self-taught learning (STL) using ERI data, named STL-detector, which is accurate with very little labeled information. In detail, STL-detector consists of two components. In the <i>unsupervised feature learning</i> component, we construct a 3-D convolutional neural networks (3-D-CNN) autoencoder trained with an amount of unlabeled data, which forms a succinct high-level input representation and significantly improve detection performance. In the <i>supervised classification</i> component, we utilize the random forest (RF) as the classifier, which is trained on very little labeled data, to detect ride-sharing cars/others. The experimental results demonstrate that our STL-detector model can detect ride-sharing cars with better performance compared with other baselines on a set of train and test samples. Furthermore, we apply our model to a real-world scenario to detect ride-sharing cars and conduct a comparative analysis on the behavior of detected ride-sharing cars and taxis.