A Framework for Ultra Low-Power Hardware Accelerators Using NNs for Embedded Time Series Classification

In embedded applications that use neural networks (NNs) for classification tasks, it is important to not only minimize the power consumption of the NN calculation, but of the whole system. Optimization approaches for individual parts exist, such as quantization of the NN or analog calculation of arithmetic operations. However, there is no holistic approach for a complete embedded system design that is generic enough in the design process to be used for different applications, but specific in the hardware implementation to waste no energy for a given application. Therefore, we present a novel framework that allows an end-to-end ASIC implementation of a low-power hardware for time series classification using NNs. This includes a neural architecture search (NAS), which optimizes the NN configuration for accuracy and energy efficiency at the same time. This optimization targets a custom designed hardware architecture that is derived from the key properties of time series classification tasks. Additionally, a hardware generation tool is used that creates a complete system from the definition of the NN. This system uses local multi-level RRAM memory as weight and bias storage to avoid external memory access. Exploiting the non-volatility of these devices, such a system can use a power-down mode to save significant energy during the data acquisition process. Detection of atrial fibrillation (AFib) in electrocardiogram (ECG) data is used as an example for evaluation of the framework. It is shown that a reduction of more than 95% of the energy consumption compared to state-of-the-art solutions is achieved.

Real-Time Malware Process Detection and Automated Process Killing

Perimeter-based detection is no longer sufficient for mitigating the threat posed by malicious software. This is evident as antivirus (AV) products are replaced by endpoint detection and response (EDR) products, the latter allowing visibility into live machine activity rather than relying on the AV to filter out malicious artefacts. This paper argues that detecting malware in real-time on an endpoint necessitates an automated response due to the rapid and destructive nature of some malware. The proposed model uses statistical filtering on top of a machine learning dynamic behavioural malware detection model in order to detect individual malicious processes on the fly and kill those which are deemed malicious. In an experiment to measure the tangible impact of this system, we find that fast-acting ransomware is prevented from corrupting 92% of files with a false positive rate of 14%. Whilst the false-positive rate currently remains too high to adopt this approach as-is, these initial results demonstrate the need for a detection model that is able to act within seconds of the malware execution beginning; a timescale that has not been addressed by previous work.

An in-process multi-feature data fusion nondestructive testing approach for wire arc additive manufacturing

Purpose The major problem that limits the widespread use of WAAM technology is the forming quality. However, most of the current research focuses on post-process detections that are time-consuming, expensive and destructive. This paper aims to achieve the on-line detection and classification of the common defects, including hump, deposition collapse, deviation, internal pore and surface slag inclusion. Design/methodology/approach This paper proposes an in-process multi-feature data fusion nondestructive testing method based on the temperature field of the WAAM process. A thermal imager is used to collect the temperature data of the deposition layer in real-time. Efficient processing methods are proposed in this paper, such as the temperature stack algorithm, width extraction algorithm and a classification model based on a residual neural network. Some features closely related to the forming quality were extracted, containing the profile image and width curve of the deposition layer and abnormal temperature features in longitudinal and cross-sections. These features are used to achieve the detection and classification of defects. Findings Thermal non-destructive testing is a potentially superior technology for in-process detection in the industrial field. Based on the temperature field, extracting the most relevant features of the defect information is crucial. This paper pushes current infrared (IR) monitoring methods toward real-time detection and proposes an in-process multi-feature data fusion non-destructive testing method based on the temperature field of the WAAM process. Originality/value In this paper, the single-layer and multi-layer WAAM samples are preset with various defects, such as hump, deposition collapse, deviation, pore and slag inclusion. A multi-feature nondestructive testing methodology is proposed to realize the in-process detection and classification of the defects. A temperature stack algorithm is proposed, which improves the detection accuracy of profile change and solves the problem of uneven temperature from arc striking to arc extinguishing. The combination of residual neural network greatly improves the accuracy and efficiency of detection.

Dense Robust 3D Reconstruction and Measurement for 3D Printing Process Based on Vision

The 3D printing process lacks real-time inspection, which is still an open-loop manufacturing process, and the molding accuracy is low. Based on the 3D reconstruction theory of machine vision, in order to meet the applicability requirements of 3D printing process detection, a matching fusion method is proposed. The fast nearest neighbor (FNN) method is used to search matching point pairs. The matching point information of FFT-SIFT algorithm based on fast Fourier transform is superimposed with the matching point information of AKAZE algorithm, and then fused to obtain more dense feature point matching information and rich edge feature information. Combining incremental SFM algorithm with global SFM algorithm, an integrated SFM sparse point cloud reconstruction method is developed. The dense point cloud is reconstructed by PMVs algorithm, the point cloud model is meshed by Delaunay triangulation, and then the accurate 3D reconstruction model is obtained by texture mapping. The experimental results show that compared with the classical SIFT algorithm, the speed of feature extraction is increased by 25.0%, the number of feature matching is increased by 72%, and the relative error of 3D reconstruction results is about 0.014%, which is close to the theoretical error.

Muonium and Water in Histidine Amino Acid

In order to apply muon spin rotation and relaxation method for study of life sciences like electron transfer process, detection of molecular concentration, photosynthesis process, etc., theoretical study to understand the stopping sites of muon and its charge species in the macromolecules is necessary. In the systematic theoretical study to know the behaviour of muon and muonium in water hydrated biological macromolecules like protein and DNA through the first-principles approach, the behaviour of a water molecule in the presence of muonium in histidine amino acid with extended main chain is presented here. The sites of a water molecule and a muonium in histidine amino acid are estimated. Two possible sites with potential energy 0.3 eV (approximately) for water molecule in the optimized structure of muonium in extended main chain histidine were estimated. Water in the sites is expected to contribute to enhance the intra- and inter-chain electron transfer in the system as reported experimentally.

Computation of distribution of relaxation times by Tikhonov regularization for Li ion batteries: usage of L-curve method

In this paper, the distribution of relaxation times (DRTs) functions are calculated numerically in Matlab for synthetic impedance data from single parallel $$RC$$ RC circuit and two parallel $$RC$$ RC circuits connected in series, experimental impedance data from supercapacitors and α-LiFeO2 anode based Li ion batteries. The quality of the impedance data is checked with the Kramers–Krönig (KK) relations. The DRTs are calculated within the KK compatible regime for all the systems using Tikhonov regularization (TR) method. Here we use a fast and simple L-curve method to estimate the TR parameter (λ) for regularization of the Fredholm integral equations of first kind in impedance. Estimation of the regularization parameters are performed effectively from the offset of the global corner of the L-curve rather than simply using the global corner. The physical significances of DRT peaks are also discussed by calculating the effective resistances and capacitances coupled with peak fitting program. For instance, two peaks in the DRTs justify the electrical double layer capacitance and ion diffusion phenomena for supercapacitors in low to intermediate frequencies respectively. Moreover, the surface film effect, Li/electrolyte and electrode/electrolyte charge transfer related processes are identified for α-LiFeO2 anode based Li-ion batteries. This estimation of the offset of the global corner extends the L-curve approach coupled with the Tikhonov regularization in the field of electrochemistry and can also be applied in similar process detection methods.

In-process detection of grinding burn using machine learning

The improvement of industrial grinding processes is driven by the objective to reduce process time and costs while maintaining required workpiece quality characteristics. One of several limiting factors is grinding burn. Usually applied techniques for workpiece burn are conducted often only for selected parts and can be time consuming. This study presents a new approach for grinding burn detection realized for each ground part under near-production conditions. Based on the in-process measurement of acoustic emission, spindle electric current, and power signals, time-frequency transforms are conducted to derive almost 900 statistical features as an input for machine learning algorithms. Using genetic programming, an optimized combination between feature selector and classifier is determined to detect grinding burn. The application of the approach results in a high classification accuracy of about 99% for the binary problem and more than 98% for the multi-classdetection case, respectively.