Search Results (26)

(1) Background: The digital traces of unmanned aerial vehicles (UAVs) are becoming increasingly important in criminal incidents, the violation of airspace and in military operations, thus making the reconstruction of the digital traces a critical task. But, current tools like DatCon, Autopsy and GRYPHON cannot recover telemetry when the flight logs are logically damaged, fragmented or partially deleted and don’t offer any quantitative measurement of the confidence of the recovered information. (2) Methods: A unified hardware-software complex, including a forensic workstation, a hardware write-blocker and SD/microSD/eMMC adapters; a set of software modules for extracting artifacts from files, structural parsing of DAT/BIN/CSV log, neural network reconstruction of missing telemetry using a two-layer LSTM architecture; a multi-source correlation module that combines flight logs, telemetry, media metadata and controller artifacts; a module, Confidence Score (CS), that computes a reliability measure in $[0,1]$; and a visualization module to generate a reconstructed trajectory on an electronic map. (3) Results: The complex has been tested on 105 flights on 10 different UAVs, 492 flight logs were gathered, 10,435 were the media item files and 624 GB was the amount of storage during acquisition. The carving stage recovers 98.7% of artifacts across the eight signature classes, the LSTM module recovers all five telemetry parameters with $R^2>0.99$ and a single-step horizontal position error of 6.8 m, which is reduced to 4.7 m after multi-source correlation (below the 5 m operational target consistent with consumer-GNSS precision); the dependence on gap length is described by the empirical growth law ${\epsilon }_{\mathrm{horiz}}\approx 4.84·G^{1.44}$ m; 46.8% of recovered records fall within the high-confidence band of $\mathrm{CS}\ge 0.8$; and the complex outperforms DatCon, Autopsy + DJI Analyzer and GRYPHON by 22–35 percentage points in end-to-end record recovery and by a factor of ∼2.6 in mean horizontal error (4.7 m vs. 12.4–18.7 m). (4) Conclusions: The combined write-blocked hardware acquisition, neural reconstruction of telemetry, and quantitative confidence index provides a forensically structured pipeline that fills an existing gap in UAV digital forensics; we note that technical reconstruction accuracy does not by itself confer legal admissibility, which remains a function of jurisdiction-specific evidentiary standards discussed in the Conclusions. Full article

Masked diffusion language models (MDLMs) apply full bidirectional attention at every denoising step, which incurs $\mathcal{O}\left(T{n}^2\right)$ cost in the number of steps T and the sequence length n. For an 8B parameter model at n = 8192 with T = 128, the KV cache alone exceeds 40 GB and rules out long document generation on a single GPU. We introduce SW-SpeedDLM, an inference wrapper for pretrained MDLMs that generates sequences of up to 16,384 tokens on one A100 with 40 GB. The framework comprises three components, each targeting a distinct bottleneck. Segmented SlidingWindow Denoising (SSWD) restricts each denoising loop to a window of W tokens and reduces the per-step cost to $\mathcal{O}(W^{2}n/S)$. Cross-Segment KV Compression (CSKV) encodes each completed window into C summary tokens that later windows attend to at $\mathcal{O}\left(C\right)$ cost. Window Level Speculative Acceptance (WLSA) lets a small draft model propose k denoising steps that the target model verifies in a single forward pass, yielding up to 2.4× per window speedup. We prove that WLSA preserves the exact marginal distribution of the target model. On MDLM-860M and LLADA-8B across PG-19, LongBench, and WritingPrompts, SW-SpeedDLM achieves 3.7× higher throughput at n = 8192 than full attention generation at n = 2048 (its maximum feasible length) and lowers peak memory by 2.5× relative to full attention at n = 4096, the longest length full attention can support before exhausting GPU memory, while also increasing PG-19 bits per character by only 0.18. Full article

Due to the Von Neumann bottleneck of traditional CMOS computing, there is an urgent need to develop in-memory logic devices with low power consumption. In this work, we demonstrate ferroelectric diode devices based on the TiN/Hf_0.5Zr_0.5O₂/HfO₂/TiN structure, implementing 16 Boolean logic operations through single-step or multi-step (2–3 steps) cascade and achieving attojoule-level one-bit full-adder computation. The TiN/Hf_0.5Zr_0.5O₂/HfO₂/TiN ferroelectric diode exhibits non-destructive readout and bidirectional rectification characteristics, with the conduction mechanism following Schottky emission behavior in the on-state. Based on its bidirectional rectification characteristics, we designed and simulated the circuit scheme of 16 Boolean logic and one-bit full-adder through cascaded operations. Both the input and output logic values are represented in the form of resistance, without the need for additional form conversion circuits. The state writing is performed by pulse-controlled polarization flipping, and the state reading is non-destructive. The logic circuits in this work demonstrate superior performance with ultralow computing power consumption in simulation. This breakthrough establishes a foundation for developing energy-efficient and scalable in-memory computing systems. Full article

The increasing demand for novel tissue engineering (TE) applications in bone tissue regeneration underscores the importance of exploring advanced manufacturing techniques and biomaterials for personalised treatment approaches. Three-dimensional printing (3DP) technology facilitates the development of implantable devices with intricate geometries, enabling patient-specific therapeutic solutions. Although Fused Filament Fabrication (FFF) and Direct Ink Writing (DIW) are widely utilised for fabricating bone-like implants, the need for multiple processing steps often prolongs the overall production time. In this study, a single-step melt-extrusion 3DP technique was performed to develop multi-material scaffolds including bioceramics, hydroxyapatite (HA), and β-tricalcium phosphate (TCP) in both their bioactive and calcined forms at 10% and 20% w/w, within polycaprolactone (PCL) matrices. Printing parameters were optimised, and physicochemical properties of all biomaterials and final forms were evaluated. Thermal degradation and surface morphology analyses assessed the consistency and distribution of the ceramics across the different formulations. The tensile testing of the scaffolds defined the impact of each ceramic type and wt% on scaffold flexibility performance, while in vitro cell studies determined the cytocompatibility efficiency. Hence, all 3D-printed PCL–ceramic composite scaffolds achieved structural integrity and physicochemical and thermal stability. The mechanical profile of extruded samples was relevant to the ceramic consistency, providing valuable insights for further mechanotransduction investigations. Notably, all materials showed high cell viability and proliferation, indicating strong biocompatibility. Therefore, this additive manufacturing (AM) process is a precise and fast approach for developing biomaterial-based scaffolds, with potential applications in surgical restoration and support of segmental bone defects. Full article

Nowadays, with the common use of smartphones in daily lives, mobile applications have become popular around the world, which will lead to a rise in Flutter framework. Developed by Google, Flutter with Dart programming provides a cross-platform development environment to create visually appealing and responsive user interfaces across mobile, web, and desktop platforms using a single codebase. However, due to time and staff limitations, the Flutter/Dart programming course is not included in curricula, even in IT departments in universities. Therefore, independent learning environments for students are essential to meet this growing popularity. Previously, we have developed programming learning assistant system (PLAS) as a web-browser-based self-learning platform for novice students. PLAS offers various types of exercise problems designed to cultivate programming skills step-by-step through a lot of code reading and code writing practices. Among them, one particular type is the code modification problem (CMP), which asks to modify the given source code to satisfy the new specifications. CMP is expected to be solved by novices with little effort if they have knowledge of other programming languages. Thus, PLAS with CMP will be an excellent platform for independent learning. In this paper, we present PLAS with CMP for the independent learning of Flutter/Dart programming. To improve the readability of the source code by students, we provided rich comments on grammar or behaviors. Besides, the code can be downloaded so that students can check and run it on an IDE. For evaluations, we generated 38 CMP instances for basic and multimedia/storage topics in Flutter/Dart programming and assigned them to 21 master students at Okayama University, Japan, who have never studied it. The results confirm the validity of the proposal. Full article

Pen-holding postures (PHPs) can significantly affect the speed and quality of writing, and incorrect postures can lead to health problems. This paper presents and experimentally implements a methodology for quickly recognizing and correcting poor writing postures using a digital dot matrix pen. The method first extracts basic handwriting information, including page number, handwriting coordinates, movement trajectory, pen tip pressure, stroke sequence, and pen handling time. This information is then used to generate writing features that are fed into our proposed fusion classification model, which combines a simple parameter-free attention module for convolutional neural networks (CNNs) called NetworkSimAM, CNNs, and an extension of the well-known long short-term memory (LTSM) called Mogrifier LSTM or MLSTM. Finally, the method ends with a classification step (Softmax) to recognize the type of PHP. The implemented method achieves significant results through receiver operating characteristic (ROC) curves and loss functions, including a recognition accuracy of 72%, which is, for example, higher than that of the single-stroke model (i.e., TabNet incorporating SimAM). The obtained results show that a promising solution is provided for accurate and efficient PHP recognition and has the potential to improve writing speed and quality while reducing health problems induced by incorrect postures. Full article

In this paper, the first- and fourth-order fiber Bragg grating (FBG)-based axial strain sensors are proposed. The FBGs are inscribed in step-index polymer optical fibers (POFs) (TOPAS core and ZEONEX cladding) via the point-by-point (PbP) direct-writing technique. A first-order FBG with a single peak is obtained with a pulse fluence of 7.16 J/cm², showing a strain sensitivity of 1.17 pm/με. After that, a fourth-order FBG with seven peaks is obtained with a pulse fluence of 1.81 J/cm² with a strain sensitivity between 1.249 pm/με and 1.296 pm/με. With a higher fluence of 2.41 J/cm², a second fourth-order FBG with five peaks is obtained, each of which is split into two peaks due to high birefringence (Hi-Bi) of ~5.4 × 10⁻⁴. The two split peaks present a strain sensitivity of ~1.44 pm/με and ~1.55 pm/με, respectively. The peak difference corresponding to Hi-Bi presents a strain sensitivity of ~0.11 pm/με and could potentially be used for simultaneous dual-parameter measurement, such as temperature and strain. Full article

Urdu is a widely spoken and narrated language in several South-Asian countries and communities worldwide. It is relatively hard to recognize Urdu text compared to other languages due to its cursive writing style. The Urdu text script belongs to a non-Latin cursive family script like Arabic, Hindi and Chinese. Urdu is written in several writing styles, among which ‘Nastaleeq’ is the most popular and widely used font style. A gap still poses a challenge for localization/detection and recognition of Urdu Nastaleeq text as it follows modified version of Arabic script. This research study presents a methodology to recognize and classify Urdu text in Nastaleeq font, regardless of the text position in the image. The proposed solution is comprised of a two-step methodology. In the first step, text detection is performed using the Connected Component Analysis (CCA) and Long Short-Term Memory Neural Network (LSTM). In the second step, a hybrid Convolution Neural Network and Recurrent Neural Network (CNN-RNN) architecture is deployed to recognize the detected text. The image containing Urdu text is binarized and segmented to produce a single-line text image fed to the hybrid CNN-RNN model, which recognizes the text and saves it in a text file. The proposed technique outperforms the existing ones by achieving an overall accuracy of 97.47%. Full article

Laser scribing is a technique that converts carbon-rich precursors into 3D-graphene nanomaterial via direct, single-step, and maskless laser writing in environmental conditions and using a scalable approach. It allows simple, fast, and reagentless production of a promising material with outstanding physicochemical features to create novel electrochemical sensors and biosensors. This review addresses different strategies for fabricating laser-scribed graphene (LSG) devices and their association with nanomaterials, polymers, and biological molecules. We provide an overview of their applications in environmental and health monitoring, food safety, and clinical diagnosis. The advantages of their integration with machine learning models to achieve low bias and enhance accuracy for data analysis is also addressed. Finally, in this review our insights into current challenges and perspectives for LSG electrochemical sensors are presented. Full article

Femtosecond direct laser writing is a well-established and robust technique for the fabrication of photonic structures. Herein, we report on the fabrication of buried waveguides in AgPO₃ silver metaphosphate glasses, as well as, on the erase and re-writing of those structures, by means of a single femtosecond laser source. Based on the fabrication procedure, the developed waveguides can be erased and readily re-inscribed upon further femtosecond irradiation under controlled conditions. Namely, for the initial waveguide writing the employed laser irradiation power was 2 J/cm² with a scanning speed of 5 mm/s and a repetition rate of 200 kHz. Upon enhancing the power to 16 J/cm² while keeping constant the scanning speed and reducing the repetition rate to 25 kHz, the so formed patterns were readily erased. Then, upon using a laser power of 2 J/cm² with a scanning speed of 1 mm/s and a repetition rate of 200 kHz the waveguide patterns were re-written inside the glass. Scanning electron microscopy (SEM) images at the cross-section of the processed glasses, combined with spatial Raman analysis revealed that the developed write/erase/re-write cycle, does not cause any structural modification to the phosphate network, rendering the fabrication process feasible for reversible optoelectronic applications. Namely, it is proposed that this non-ablative phenomenon lies on the local relaxation of the glass network caused by the heat deposited upon pulsed laser irradiation. The resulted waveguide patterns Our findings pave the way towards new photonic applications involving infinite cycles of write/erase/re-write processes without the need of intermediate steps of typical thermal annealing treatments. Full article

In recent years, the scaling down that Moore’s Law relies on has been gradually slowing down, and the traditional von Neumann architecture has been limiting the improvement of computing power. Thus, neuromorphic in-memory computing hardware has been proposed and is becoming a promising alternative. However, there is still a long way to make it possible, and one of the problems is to provide an efficient, reliable, and achievable neural network for hardware implementation. In this paper, we proposed a two-layer fully connected spiking neural network based on binary MRAM (Magneto-resistive Random Access Memory) synapses with low hardware cost. First, the network used an array of multiple binary MRAM cells to store multi-bit fixed-point weight values. This helps to simplify the read/write circuit. Second, we used different kinds of spike encoders that ensure the sparsity of input spikes, to reduce the complexity of peripheral circuits, such as sense amplifiers. Third, we designed a single-step learning rule, which fit well with the fixed-point binary weights. Fourth, we replaced the traditional exponential Leak-Integrate-Fire (LIF) neuron model to avoid the massive cost of exponential circuits. The simulation results showed that, compared to other similar works, our SNN with 1184 neurons and 313,600 synapses achieved an accuracy of up to 90.6% in the MNIST recognition task with full-resolution (28 × 28) and full-bit-depth (8-bit) images. In the case of low-resolution (16 × 16) and black-white (1-bit) images, the smaller version of our network with 384 neurons and 32,768 synapses still maintained an accuracy of about 77%, extending its application to ultra-low-cost situations. Both versions need less than 30,000 samples to reach convergence, which is a >50% reduction compared to other similar networks. As for robustness, it is immune to the fluctuation of MRAM cell resistance. Full article

This paper reports on single step and rapid fabrication of interdigitated electrodes (IDEs) using an inkjet printing-based approach. A commercial inkjet-printed circuit board (PCB) printer was used to fabricate the IDEs on a glass substrate. The inkjet printer was optimized for printing IDEs on a glass substrate using a carbon ink with a specified viscosity. Electrochemical impedance spectroscopy in the frequency range of 1 Hz to 1 MHz was employed for chemical sensing applications using an electrochemical workstation. The IDE sensors demonstrated good nitrite quantification abilities, detecting a low concentration of 1 ppm. Taste simulating chemicals were used to experimentally analyze the ability of the developed sensor to detect and quantify tastes as perceived by humans. The performance of the inkjet-printed IDE sensor was compared with that of the IDEs fabricated using maskless direct laser writing (DLW)-based photolithography. The DLW–photolithography-based fabrication approach produces IDE sensors with excellent geometric tolerances and better sensing performance. However, inkjet printing provides IDE sensors at a fraction of the cost and time. The inkjet printing-based IDE sensor, fabricated in under 2 min and costing less than USD 0.3, can be adapted as a suitable IDE sensor with rapid and scalable fabrication process capabilities. Full article

This work reports the development and optimization of a rapid and low-cost pen-on-paper plotting approach for the fabrication of paper-based analytical devices (PADs) using commercial writing stationery. The desired fluidic patterns were drawn on the paper substrate with commercial marker pens using an inexpensive computer-controlled x–y plotter. For the fabrication of electrochemical PADs, electrodes were further deposited on the devices using a second x–y plotting step with commercial writing pencils. The effect of the fabrication parameters (type of paper, type of marker pen, type of pencil, plotting speed, number of passes, single- vs. double-sided plotting), the chemical resistance of the plotted devices to different solvents and the structural rigidity to multiple loading cycles were assessed. The analytical utility of these devices is demonstrated through application in optical sensing of total phenols using reflectance calorimetry and in electrochemical sensing of paracetamol and ascorbic acid. The proposed manufacturing approach is simple, low cost, flexible, rapid and fit-for-purpose and enables the fabrication of sub-“one-dollar” PADs with satisfactory mechanical and chemical resistance and good analytical performance. Full article

The application of nanopatterned electrode materials is a promising method to improve the performance of thin-film optoelectronic devices such as organic light-emitting diodes (OLEDs) and organic photovoltaics. Light coupling to active layers is enhanced by employing nanopatterns specifically tailored to the device structure. A range of different nanopatterns is typically evaluated during the development process. Fabrication of each of these nanopatterns using electron-beam lithography is time- and cost-intensive, particularly for larger-scale devices, due to the serial nature of electron beam writing. Here, we present a method to generate nanopatterns of varying depth with different nanostructure designs from a single one-dimensional grating template structure with fixed grating depth. We employ multiple subsequent steps of UV nanoimprint lithography, curing, and ion beam etching to fabricate greyscale two-dimensional nanopatterns. In this work, we present variable greyscale nanopatterning of the widely used electrode material indium tin oxide. We demonstrate the fabrication of periodic pillar-like nanostructures with different period lengths and heights in the two grating directions. The patterned films can be used either for immediate device fabrication or pattern reproduction by conventional nanoimprint lithography. Pattern reproduction is particularly interesting for the large-scale, cost-efficient fabrication of flexible optoelectronic devices. Full article

Nowadays, the demand for surface functionalized plastics is constantly rising. To address this demand with an industry compatible solution, here a strategy is developed for producing hierarchical microstructures on polyethylene terephthalate (PET) by hot embossing using a stainless steel stamp. The master was structured using three laser-based processing steps. First, a nanosecond-Direct Laser Writing (DLW) system was used to pattern dimples with a depth of up to 8 µm. Next, the surface was smoothed by a remelting process with a high-speed laser scanning at low laser fluence. In the third step, Direct Laser Interference Patterning (DLIP) was utilized using four interfering sub-beams to texture a hole-like substructure with a spatial period of 3.1 µm and a depth up to 2 µm. The produced stamp was used to imprint PET foils under controlled temperature and pressure. Optical confocal microscopy and scanning electron microscopy imaging showed that the hierarchical textures could be accurately transferred to the polymer. Finally, the wettability of the single- and multi-scaled textured PET surfaces was characterized with a drop shape analyzer, revealing that the highest water contact angles were reached for the hierarchical patterns. Particularly, this angle was increased from 77° on the untreated PET up to 105° for a hierarchical structure processed with a DLW spot distance of 60 µm and with 10 pulses for the DLIP treatment. Full article