Real-time Computation Research Articles

Earliest Deadline First Scheduling for Real-Time Computing in Sustainable Sensors

Energy harvesting is a green technology that authorizes small electronic devices to be supplied for perpetual operation. It enables wireless sensors to be integrated in applications that previously were not feasible with conventional battery-powered designs. Intermittent computing and scheduling are the two central aspects of designing a Real-Time Energy Harvesting (RTEH) sensor, generally used to monitor a mission critical process. Traditional scheduling algorithms fail to timely execute the hard deadline tasks because they accommodate no fluctuations in power supply and therefore no intermittent computing. A suitable energy-harvesting-aware scheduling algorithm has been proposed so as to achieve a higher schedulability rate. Unlike the classical EDF (Earliest Deadline First) scheduler, the ED-H algorithm is idling and clairvoyant, with an improved performance in terms of the deadline missing ratio. This paper reviews the main advances in dynamic priority scheduling based on EDF for energy-neutral systems.

Safety, Feasibility and Efficacy of Lokomat® and Armeo®Spring Training in Deconditioned Paediatric, Adolescent and Young Adult Cancer Patients

Rehabilitation robotics is a field of study which aims to understand and augment rehabilitation through the use of robotics devices. This proof of concept study aimed to test the safety (no. adverse events, incidence of infection), feasibility (program demand, adherence, participant satisfaction) and efficacy (Peak Oxygen uptake (VO2peak), 6-min walk test, gait speeds, Canadian Occupational Performance Measure, quality of life) of Lokomat® and Armeo®Spring training in children and adolescents and young adults (AYAs) during or soon after cancer treatment. This was a 6-week single arm pre-post study. Cancer patients with significant musculoskeletal, neurological, gait and/or upper limb deficiency aged 5 to 25 years were recruited. The rehabilitation program included access to two robotic orthoses: the Lokomat® and/or Armeo®Spring. Robotic devices utilised real-time biofeedback and computer games to engage and guide participants through a repetitive functional range of movement aimed at improving functional deficiencies. Progressive increases in exercise intensity and duration were encouraged. Twentey-eight participants were approached for study; twenty-one consented. Seventy-six percent completed the six-week intervention with an overall adherence of 83%. The mean participant satisfaction score was 8.8/10. Forty-nine adverse events were recorded throughout the course of the study, forty-five grade 1, three grade 2 and one grade 3. No adverse events led to withdrawal from the study. Preliminary efficacy results indicate large beneficial effects on VO2peak (r = 0.63), 10 m comfortable pace walk (r = 0.51) and maximal pace walk (r = 0.60), 6-min walk test (r = 0.60), maximal back and leg strength (r = 0.71), trunk flexibility (r = 0.60), The European Organization for the Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ C30) (r = 0.61), Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT F) r = 0.53 and the Canadian Occupational Performance Measure, satisfaction (r = 0.88) and performance scores (r = 0.83), and moderate beneficial effects on Leisure Score Index (LSI) (r = 0.30). Our results suggest that Lokomat® and Armeo®Spring training is safe and feasible for use in children and AYAs who are currently undergoing or have recently completed cancer therapy. A larger controlled trial investigating the efficacy of robotics rehabilitation in this cohort is warranted.

A Low-Power Hardware Architecture for Real-Time CNN Computing.

Convolutional neural network (CNN) is widely deployed on edge devices, performing tasks such as objective detection, image recognition and acoustic recognition. However, the limited resources and strict power constraints of edge devices pose a great challenge to applying the computationally intensive CNN models. In addition, for the edge applications with real-time requirements, such as real-time computing (RTC) systems, the computations need to be completed considering the required timing constraint, so it is more difficult to trade off between computational latency and power consumption. In this paper, we propose a low-power CNN accelerator for edge inference of RTC systems, where the computations are operated in a column-wise manner, to realize an immediate computation for the currently available input data. We observe that most computations of some CNN kernels in deep layers can be completed in multiple cycles, while not affecting the overall computational latency. Thus, we present a multi-cycle scheme to conduct the column-wise convolutional operations to reduce the hardware resource and power consumption. We present hardware architecture for the multi-cycle scheme as a domain-specific CNN architecture, which is then implemented in a 65 nm technology. We prove our proposed approach realizes up to 8.45%, 49.41% and 50.64% power reductions for LeNet, AlexNet and VGG16, respectively. The experimental results show that our approach tends to cause a larger power reduction for the CNN models with greater depth, larger kernels and more channels.

Inexpensive High-Throughput Multiplexed Biomarker Detection Using Enzymatic Metallization with Cellphone-Based Computer Vision.

Multiplexed biomarker detection can play a critical role in reliable and comprehensive disease diagnosis and prediction of outcome. Enzyme-linked immunosorbent assay (ELISA) is the gold standard method for immunobinding-based biomarker detection. However, this is currently expensive, limited to centralized laboratories, and usually limited to the detection of a single biomarker at a time. We present a low-cost, smartphone-based portable biosensing platform for high-throughput, multiplexed, sensitive, and quantitative detection of biomarkers from single, low-volume drops (<1 μL) of clinical samples. Biomarker binding to spotted capture antigens is converted, via enzymatic metallization, to the localized surface deposition of amplified, dry-stable, silver metal spots whose darkness is proportional to biomarker concentration. A custom smartphone application is developed, which uses real-time computer vision to enable easy optical detection of the deposited metal spots and sensitive and reproducible quantification of the biomarkers. We demonstrate the use of this platform for high-throughput, multiplexed detection of multiple viral antigen-specific antibodies from convalescent COVID-19 patient serum as well as vaccine-elicited antibody responses from uninfected vaccine-recipient serum and show that distinct multiplexed antibody fingerprints are observed among them.

Optical-electronic hybrid Fourier convolutional neural network based on super-pixel complex-valued modulation.

An optical-electronic hybrid convolutional neural network (CNN) system is proposed and investigated for its parallel processing capability and system design robustness. It is regarded as a practical way to implement real-time optical computing. In this paper, we propose a complex-valued modulation method based on an amplitude-only liquid-crystal-on-silicon spatial light modulator and a fixed four-level diffractive optical element. A comparison of computational results of convolutions between different modulation methods in the Fourier plane shows the feasibility of the proposed complex-valued modulation method. A hybrid CNN model with one convolutional layer of multiple channels is proposed and trained electrically for different classification tasks. Our simulation results show that this model has a classification accuracy of 97.55% for MNIST, 88.81% for Fashion MNIST, and 56.16% for Cifar10, which outperforms models using only amplitude or phase modulation and is comparable to the ideal complex-valued modulation method.

Crankshaft Stress Telemetry Device and Its Low Power Strategy

In view of the fact that the crankshaft of an internal combustion engine lacks the support of measured stress data due to the closed working environment, a real-time stress test device for crankshaft and its low power strategy were studied. An embedded telemetry technology is used to solve the system miniaturization and low-power strategy. The integrated circuit chip is used to realize the miniaturization design, and a specialized power management scheme is designed to realize the low-power strategy. The actual test shows that the device can run continuously and stably under the environment of high temperature and oil pollution inside the crankcase and can record the stress curve of the crankshaft and transmit it to the remote computer in real time, which realizes the data visualization on the computer. The designed low-power strategy can help the device complete a single test period, and the device has higher promotion value and practical significance.

Ultra-Precise High-Speed Untethered Manipulation of Magnetic Scanning Microprobe in Aqueous Solutions

This article describes the design and development of an integrated system that renders ultra-precise and high-speed, untethered manipulation of a single magnetic scanning microprobe in aqueous solutions under a microscope. The system uses a six-input-three-output hexapole electromagnetic actuator to control the magnetic gradient force exerted on the probe and a three-dimensional vision-based motion tracking system to enable feedback control. The control system has three core functions that enables superior manipulation of the magnetic microprobe. The six-input-six-output closed-loop magnetic flux control significantly suppresses the effect of magnetic hysteresis and greatly increases the bandwidth of magnetic flux generation. The optimal flux allocation solves four issues in multipole magnetic force generation: redundancy; coupling; nonlinearity; and position-dependency. The position-dependent discrete-time motion control law determines the required force for stabilization and tracking. These three core functions have been implemented using a high-speed field programmable gate array (FPGA) system. Experiments have verified that the use of digital control technology, high-speed electronics, mathematical modeling, and real-time computation has achieved superior manipulation of a single magnetic microprobe in aqueous solutions.

Comparison of Processors for Real-time Homography Computation

Comparison of Processors for Real-time Homography Computation

Efficient Distributed Threshold-Based Offloading for Large-Scale Mobile Cloud Computing

Mobile cloud computing enables compute-limited mobile devices to perform real-time intensive computations such as speech recognition or object detection by leveraging powerful cloud servers. An important problem in large-scale mobile cloud computing is computational offloading, where each mobile device decides when and how much computation should be uploaded to cloud servers by considering the local processing delay and the cost of using cloud servers. In this paper, we develop a distributed threshold-based offloading algorithm where it uploads an incoming computing task to cloud servers if the number of tasks queued at the device reaches the threshold and processes it locally otherwise. The threshold is updated iteratively based on the computational load and the cost of using cloud servers. We formulate the problem as a symmetric game, and characterize the sufficient and necessary conditions for the existence and uniqueness of the Nash Equilibrium (NE) assuming exponential service times. Then, we show the convergence of our proposed distributed algorithm to the NE when the NE exists. Further, we characterize the performance gap between cost under our proposed distributed algorithm and the minimum cost in terms of Price of Anarchy (PoA) when the cost of using cloud servers is high. Finally, we perform extensive simulations to validate our theoretical findings, demonstrate the efficiency of our proposed distributed algorithm under various scenarios such as hyperexponential service times, imperfect server utilization estimation, and asynchronous threshold updates, and reveal the superior performance of threshold-based policies over their probabilistic counterpart.

Hierarchical Motion Planning for Autonomous Vehicles in Unstructured Dynamic Environments

This letter presents a hierarchical motion planner for generating smooth and feasible trajectories for autonomous vehicles in unstructured environments with static and moving obstacles. The framework enables real-time computation by progressively shrinking the solution space. First, a graph searcher based on combined heuristic and partial motion planning is proposed for finding coarse trajectories in spatiotemporal space. To enable fast online planning, a time interval-based algorithm that considers obstacle prediction trajectories is proposed, which uses line segment intersection detection to check for collisions. Second, to practically smooth the coarse trajectory, a continuous optimizer is implemented in three layers, corresponding to the whole path, the near-future path and the speed profile. We use discrete points to represent the far-future path and parametric curves to represent the near-future path and the whole speed profile. The approach is validated in both simulations and real-world off-road environments based on representative scenarios, including the “wait and go” scenario. The experimental results show that the proposed method improves the success rate and travel efficiency while actively avoiding static and moving obstacles.

Joint Service Quality Control and Resource Allocation for Service Reliability Maximization in Edge Computing

Edge computing is a commonly used paradigm for providing low-latency computation services by locally deploying computation and storage resources close to the user equipments (UEs). Since the computation resource demand of the offloaded tasks of a UE is naturally a random variable, it is possible that the real-time computation capacity demand of a resource-limited hosting virtual machine (VM) or edge computing server (ECS) is larger than its computation capacity, causing unexpected delay or delay-jitter to the services, which should be avoided if possible, for delay-sensitive applications. We consider an edge computing scenario wherein the transmission links are unmanageable and computation resource demands of VM servers are stochastic. We propose a novel Logistic function-based service reliability probability (SRP) estimation model without specifying the distributions of the resource demands. We study the average SRP maximization problem (ASRPMP) in a VM-based edge computing server (ECS) by jointly optimizing the service quality ratios (SQRs) and the computation resource allocations, and we propose an alternative optimization algorithm (AOA) by decomposing the problem into a resource allocation problem (RAP) and a service quality control problem (SQCP). Based on the derived analytical solutions of the two subproblems, we propose an effective and low-complexity heuristic AOA (HAOA) to solve the ASRPMP. The simulation results obtained from both synthetic Gaussian workload data and PlanetLab trace data demonstrate that, given the same target SQR or computation resource, the proposed method can achieve similar performance compared with the convex AOA (CAOA) method with much higher complexity, and can improve the reliability of the services compared with the baseline weighted allocation method (WAM) in both high and low SRP regimes.

Receding Horizon-Based Infotaxis With Random Sampling for Source Search and Estimation in Complex Environments

This article proposes a receding horizon-based information-theoretic source search and estimation strategy for a mobile sensor in an urban environment in which an invisible harmful substance is released into the atmosphere. The mobile sensor estimates the source term including its location and release rate by using sensor observations based on Bayesian inference. The sampling-based sequential Monte Carlo method, particle filter, is employed to estimate the source term state in a highly nonlinear and stochastic system. Infotaxis, the information-theoretic gradient-free search strategy is modified to find the optimal search path that maximizes the reduction of the entropy of the source term distribution. In particular, receding horizon Infotaxis (RHI) is introduced to avoid falling into the local optima and to find more successful information gathering paths in obstacle-rich urban environments. Besides, a random sampling method is introduced to reduce the computational load of the RHI for real-time computation. The random sampling method samples the predicted future measurements based on current estimation of the source term and computes the optimal search path using sampled measurements rather than considering all possible future measurements. To demonstrate the benefit of the proposed approach, comprehensive numerical simulations are performed for various conditions. The proposed algorithm increases the success rate by about 30% and reduces the mean search time by about 40% compared with the existing information-theoretic search strategy.

A Real-Time Computer Vision Based Approach to Detection and Classification of Traffic Incidents

To constructively ameliorate and enhance traffic safety measures in Saudi Arabia, a prolific number of AI (Artificial Intelligence) traffic surveillance technologies have emerged, including Saher, throughout the past years. However, rapidly detecting a vehicle incident can play a cardinal role in ameliorating the response speed of incident management, which in turn minimizes road injuries that have been induced by the accident’s occurrence. To attain a permeating effect in increasing the entailed demand for road traffic security and safety, this paper presents a real-time traffic incident detection and alert system that is based on a computer vision approach. The proposed framework consists of three models, each of which is integrated within a prototype interface to fully visualize the system’s overall architecture. To begin, the vehicle detection and tracking model utilized the YOLOv5 object detector with the DeepSORT tracker to detect and track the vehicles’ movements by allocating a unique identification number (ID) to each vehicle. This model attained a mean average precision (mAP) of 99.2%. Second, a traffic accident and severity classification model attained a mAP of 83.3% while utilizing the YOLOv5 algorithm to accurately detect and classify an accident’s severity level, sending an immediate alert message to the nearest hospital if a severe accident has taken place. Finally, the ResNet152 algorithm was utilized to detect the ignition of a fire following the accident’s occurrence; this model achieved an accuracy rate of 98.9%, with an automated alert being sent to the fire station if this perilous event occurred. This study employed an innovative parallel computing technique for reducing the overall complexity and inference time of the AI-based system to run the proposed system in a concurrent and parallel manner.

Solving Time-Fractional reaction–diffusion systems through a tensor-based parallel algorithm

Machine Learning (ML) approach is a discussed research topic because of its benefit in several research fields. The most important issues in the training process of ML are accuracy and speed: a suitable mathematical model is critical and a fast data processing is mandatory. Fractional Calculus is involved in a large number of important applications and, recently, many ML algorithms, in order to improve accuracy of results when performing training in solving optimization problems, are based on decision and control performed by means of time-fractional models to better understand complex systems. However, the high computational cost, which characterizes the numerical solution, of this approach might be a problem for large scale Machine Learning systems. High Performance Computing (HPC) is the way of addressing the need of real time computation. In fact, through tensor-based parallel strategies designed for modern parallel architectures, Fractional Calculus tools are very helpful for the ML training step. In this contest, we consider a time-fractional diffusion system and, after introducing a suitable modification of a numerical model to solve it, we propose a related and novel parallel implementation on GPUs (Graphics Processing Units). Experiments show the gain of performance in terms of execution time and accuracy of our parallel implementation.

Concrete mix design using Java Eclipse IDE

Concrete mix design is the process of selecting the proper components and figuring out the proper ratios to achieve the desired workability, strength, and durability. As a result, an attempt has been made to create a Java program for concrete mix design by IS 10262:2019 and IS 456:2000. Computer software is created in this way to proportion input variables to a set of output variables, which avoids time-consuming computations in real time. The produced program was carefully checked. The results were compared to make sure they were accurate after being put through a series of manually check issues. The concrete mix design produced by the Java program was exact, time-saving, and computationally efficient.

Online and Real-Time Trajectory Generation Method for Unforeseen Events Using a Modified Spline Approach

This letter focuses on the problem of online and instantaneous generation of smooth trajectories and their analytical time derivatives, where future reference signals provided by a trajectory planner or a user are a-priori unknown. For this purpose, we propose a modified cubic spline polynomial which transforms the instantaneous piecewise references into smooth and continuous functions for digital control systems. The resulting reference signals are not only smooth and continuous in time but also analytically differentiable. The inferred derivative values could then be used in a variety of linear and non-linear controllers such as, Proportional-Integral-Derivative (PID), sliding mode, and backstepping. We also provide optimum and boundary values for the parameters used in the proposed technique. In addition, we present a lightweight implementation (a few lines of code) of the algorithm by which real-time computations are rapidly performed using a microcontroller. Both simulation and real experiments demonstrate that for the plants having different degrees of complexities, the proposed approach with a PD / PID controller outperforms the conventional filtered derivative-based one in terms of overshoot in the step response, robustness against noise, and trajectory tracking error.

CDBA: a novel multi-branch feature fusion model for EEG-based emotion recognition.

EEG-based emotion recognition through artificial intelligence is one of the major areas of biomedical and machine learning, which plays a key role in understanding brain activity and developing decision-making systems. However, the traditional EEG-based emotion recognition is a single feature input mode, which cannot obtain multiple feature information, and cannot meet the requirements of intelligent and high real-time brain computer interface. And because the EEG signal is nonlinear, the traditional methods of time domain or frequency domain are not suitable. In this paper, a CNN-DSC-Bi-LSTM-Attention (CDBA) model based on EEG signals for automatic emotion recognition is presented, which contains three feature-extracted channels. The normalized EEG signals are used as an input, the feature of which is extracted by multi-branching and then concatenated, and each channel feature weight is assigned through the attention mechanism layer. Finally, Softmax was used to classify EEG signals. To evaluate the performance of the proposed CDBA model, experiments were performed on SEED and DREAMER datasets, separately. The validation experimental results show that the proposed CDBA model is effective in classifying EEG emotions. For triple-category (positive, neutral and negative) and four-category (happiness, sadness, fear and neutrality), the classification accuracies were respectively 99.44% and 99.99% on SEED datasets. For five classification (Valence 1-Valence 5) on DREAMER datasets, the accuracy is 84.49%. To further verify and evaluate the model accuracy and credibility, the multi-classification experiments based on ten-fold cross-validation were conducted, the elevation indexes of which are all higher than other models. The results show that the multi-branch feature fusion deep learning model based on attention mechanism has strong fitting and generalization ability and can solve nonlinear modeling problems, so it is an effective emotion recognition method. Therefore, it is helpful to the diagnosis and treatment of nervous system diseases, and it is expected to be applied to emotion-based brain computer interface systems.

On the Design of an Algorithm for Mobility Management Over Fog Geographical Allocation for Healthcare Applications

Healthcare applications need an immediate response with minimal latency. Fog computing provides real-time monitoring and computation at the network edge so the data collected by Internet of Things (IoT) devices can be analyzed with minimal cost in real-time. However, dealing with end-user mobility becomes even more challenging with fog-assisted IoT healthcare frameworks because low latency in mobility support and fog node locating is a vital prerequisite. In this paper, we propose a fog node (FN) allocation and mobility-adapted fog computing architecture. The architecture is based on an algorithm that offers low latency for location-sensitive operations and service stability with high data broadcast quality for user mobility in fog computing. In our proposed architecture, it is no longer obligatory to either carry out a job on, or redirect it to, the cloud but instead connect with other FNs to continue with a request for a new job. The proposed architecture minimizes average network lateness by 40%–50%, bandwidth by 30%–40%, and end-to-end communication by 35%–55%. Additionally, the utilization of the FN computational resources is maximized.

Accurate TCP Position and Orientation Trajectory Generation in 6DOF Robotic Manipulators and CNC Machine Tools using FIR Filtering and Haversine Synchronisation

The use of robotics in manufacturing is rapidly growing. One promising application in particular is the use of robots as machining platforms. Robotic machining offers many advantages such as large operating envelopes, flexibility in operations and low installation and running costs. Robotic machining platforms much like computer numerically controlled (CNC) machine tools are commanded by numerical controllers (NCs) using the well established G-code part programming language (ISO-6983). Toolpaths are defined in the work piece coordinate system (WCS) as a series of discrete tool centre point (TCP) positions and tool orientations. The role of the NC is to generate smooth feed drive or robot joint commands such that the tool or end effector travels along the commanded toolpath whilst satisfying TCP position and orientation tolerances and robot/machine tool kinematic constraints. Interpolating tool orientations directly in the WCS requires complex real-time spherical interpolation computations. Overcoming this challenge, this paper introduces a C4 continuous tool path trajectory generation method for 6DOF motion. The tool centre point position is smoothed in the WCS, however, the tool orientation is smoothed directly in the rotary coordinate system without the requirement for spherical interpolation. The interpolated tool orientation trajectory follows the path of the great circle arc using the modified Haversine method eliminating non-linear interpolation errors present in linear interpolation of rotary axis positions (machine coordinate system interpolation). The proposed method is validated by simulation of a 6DOF robotic manipulator.

Solution-Processed Hybrid Europium (II) Iodide Scintillator for Sensitive X-Ray Detection.

Lead halide perovskite nanocrystals have recently demonstrated great potential as x-ray scintillators, yet they still suffer toxicity issues, inferior light yield (LY) caused by severe self-absorption. Nontoxic bivalent europium ions (Eu2+) with intrinsically efficient and self-absorption-free d-f transition are a prospective replacement for the toxic Pb2+. Here, we demonstrated solution-processed organic-inorganic hybrid halide BA10EuI12 (BA denotes C4H9NH4+) single crystals for the first time. BA10EuI12 was crystallized in a monoclinic space group of P21/c, with photoactive sites of [EuI6]4- octahedra isolated by BA+ cations, which exhibited high photoluminescence quantum yield of 72.5% and large Stokes shift of 97 nm. These properties enable an appreciable LY value of 79.6% of LYSO (equivalent to ~27,000 photons per MeV) for BA10EuI12. Moreover, BA10EuI12 shows a short excited-state lifetime (151 ns) due to the parity-allowed d-f transition, which boosts the potential of BA10EuI12 for use in real-time dynamic imaging and computer tomography applications. In addition, BA10EuI12 demonstrates a decent linear scintillation response ranging from 9.21 μGyair s-1 to 145 μGyair s-1 and a detection limit as low as 5.83 nGyair s-1. The x-ray imaging measurement was performed using BA10EuI12 polystyrene (PS) composite film as a scintillation screen, which exhibited clear images of objects under x-ray irradiation. The spatial resolution was determined to be 8.95 lp mm-1 at modulation transfer function = 0.2 for BA10EuI12/PS composite scintillation screen. We anticipate that this work will stimulate the exploration of d-f transition lanthanide metal halides for sensitive x-ray scintillators.