Search

Doors are part of the building infrastructure that mobile robots have to pass through to reach zones on the other side. If robots were to clear these obstacles, they would require human assistance, advanced end-effectors, and complex control systems, making it challenging for robots. Therefore, a robot deployed in an environment should be capable of minimizing the passing through doors as well as path distance to improve overall efficiency. This paper proposes a novel Door-Density-Aware (DDA) path planning method. A vision-based door-detecting framework based on YOLOv8 has been developed to tag the door locations in a robot’s navigation map. The proposed DDA path planner uses a door-tagged map to plan an efficient path considering the cost of moving through doors and the path distance. Genetic Algorithm (GA) and Gray Wolf Optimization (GWO) have been considered for solving this optimization problem. According to the experimental results, the proposed method can effectively detect and tag doors in the navigation map and plan efficient paths. In summary, the proposed DDA path planner with GA outperformed other approaches, achieving cost reductions of 66%, 34%, 49%, and 60% compared to random selection, DDA with GWO, GA minimizing only distance, and GWO minimizing only distance, respectively.

This review paper provides a comprehensive overview of recent advances in nuclear quadrupole resonance (NQR) spectroscopy for substance detection, highlighting its principles, methodologies, and applications. The paper elucidates the fundamental physics underlying NQR spectroscopy, emphasizing the interaction between nuclear quadrupole moments and electric field gradients. It explores the various experimental techniques and instrumentation developments that have enabled the sensitive detection and precise characterization of substances containing quadrupolar nuclei. A significant portion of the survey is dedicated to discussing the diverse applications of NQR spectroscopy, including the detection of explosives, drug pharmaceuticals, and material authentication. Furthermore, the survey examines the challenges and limitations associated with NQR spectroscopy, including issues related to signal-to-noise ratio (SNR), temperature dependency, and substance restrictions. Strategies to overcome these challenges are discussed, offering insights into the future directions of NQR spectroscopy research that includes artificial intelligence (AI), internet of things (IoT) integration, incorporating a cloud database for NQR parameter storage, and multi-modal analysis.

Stock market forecasting involves predicting fluctuations and trends in the value of financial assets, utilizing statistical and machine learning models to analyze historical market data for insights into future behavior. This practice aids investors, traders, financial institutions, and governments in making informed decisions, managing risks, and assessing economic conditions. Forecasting financial markets is difficult due to the intricate interplay of global economics, politics, and investor sentiment, making it inherently unpredictable. This study introduces a Deep Learning based Expert Framework for Stock Market forecasting (Portfolio prediction) called DLEF-SM. The methodology begins with an improved jellyfish-induced filtering (IJF-F) technique for preprocessing, effectively analyzing raw data and eliminating artifacts. To address imbalanced data and enhance data quality, pre-trained convolutional neural network (CNN) architectures, VGGFace2 and ResNet-50, are used for feature extraction. Additionally, an improved black widow optimization (IBWO) algorithm is designed for feature selection, reducing data dimensionality and preventing under-fitting. For precise stock market predictions, integrate deep reinforcement learning with artificial neural network (DRL-ANN) is proposed. Simulation outcomes reveal that the proposed framework achieves maximum forecasting accuracy, reaching 99.562%, 98.235%, and 98.825% for S&P500-S, S&P500-L, and DAX markets, respectively.

Mobile robots have emerged as versatile substitutes for human labor across diverse domains, offering promising applications in surveillance, healthcare, and beyond. Fundamental to their autonomy are the core capabilities of movement, perception, cognition, and navigation. This research introduces a novel approach known as the Dual PID based low-cost navigation system (DPLNS), designed specifically for indoor warehouse-like environments. The primary objective of this technique is to enable seamless point-to-point traversal. This is achieved through a fusion of gyroscope correction and visual PID control mechanisms. Leveraging a strategically positioned eagle-eye perspective camera, the system gained crucial insights for navigation. To ensure the uninterrupted execution of planned trajectories, the system employs the Message Queuing Telemetry Transport (MQTT) protocol. This technology ensures smooth communication and coordination of actions. The experimental validation of the proposed strategy highlights its efficacy, positioning it as a promising solution for modern warehouse automation needs. By advancing the capabilities of mobile robots within warehouse contexts, this research not only addresses immediate operational demands, but also paves the way for broader applications across various other fields. The developed system enables robust detection and successful path traversal to the designated spot through a position control operating at average of 12.3 rad/s.

Computational drug repurposing is an efficient method to utilize existing knowledge for understanding and predicting their effect on neurological diseases. The ability of a molecule to cross the blood-brain barrier is a primary criteria for effective therapy. Thus, accurate predictions by employing Machine learning models can effectively identify the drug candidates that could be repurposed for neurological conditions. This study comprehensively analyzes the performance of the well-known machine learning models on two different datasets to overcome dataset-related biases. We found that random forest and extratrees (i.e., tree-based ensembled models) have the highest accuracy with mol2vec fingerprint for BBB permeability prediction, attaining AUC_ROC of 0.9453 and 0.9601 on BBB and B3DB dataset, respectively. Additionally, we have analyzed the impact of the data balancing technique (i.e., SMOTE) to improve the specificity of the models. Finally, we have explored the impact of different fingerprint combinations on accuracy. By employing SMOTE and fingerprint combination, SVC attains the highest AUC_ROC of 0.9511 on BBB dataset. Finally, we used the best-performing models of the B3DB dataset to evaluate the BBB permeability for drugs intended to be used for repurposing. Model validation for repurposing predicted the non-passage for most antihypertensive drugs and passage for CYP17A1 cancer drugs.

Owing to limited drug testing possibilities in pregnant population, the development of computational algorithms is crucial to predict the fate of drugs in the placental barrier; it could serve as an alternative to animal testing. The ability of a molecule to effectively cross the placental barrier and reach the fetus determines the drug’s toxicological effects on the fetus. In this regard, our study aims to predict the permeability of molecules across the placental barrier. Based on publicly available datasets, several machine learning models are comprehensively analysed across different fingerprints and toolkits to find the best suitable models. Several dataset analysis models are utilised to study the data diversity. Further, this study demonstrates the application of neural network-based models to effectively predict the permeability. K-nearest neighbour (KNN), standard vector classifier (SVC) and Multi-layer perceptron (MLP) are found to be the best-performing models with a prediction percentage of 82%, 86.4% and 90.8%, respectively. Different models are compared to predict the chosen set of drugs, drugs like Aliskiren, some insulin secretagogues and glucocorticoids are found to be negative while predicting the permeability.

Touch screen devices have become ubiquitous in modern day-to-day lives. While smaller touchscreen devices provide enough user engagement with meaningful haptic feedback, large touch devices still lack meaningful haptic stimulation. Existing literature for large touch surface vibrotactile localization uses many actuators and conventional boundary conditions. Previous numerical studies on haptic localization do not address multi-frequency excitation. This research proposes a new spring-damper boundary condition for a large bar-type display. Subsequently, a mechanical model of the large touch bar surface with multiple-electrostatic vibration actuators is developed using material damping information with multi-frequency excitation. Simultaneously, an experimental setup is developed for validation of the developed model. An optimization technique to localize vibrotactile haptic rendering at multiple selected zones of the touch bar is proposed. It has been established that by varying frequencies of excitation of two electrostatic resonant actuators, localizable vibrotactile feedback can be generated across the length of the touch bar. The experimental results corroborate the simulation results. Finally, the proposed optimization strategy for multi-zone vibrotactile rendering is experimentally verified.

In this paper, the physical layer security (PLS) analysis for a decode-and-forward (DF) protocol based multi-hop hybrid radio frequency (RF)/ free space optics (FSO) is presented. Herein, two different scenarios are considered for detecting the secured information at the destination. In the first scenario, at each hop, the received signals with higher secrecy capacity are selected. However, in the second scenario, the RF and FSO signals are simply decoded and forwarded to the next hop and selection is done at the last hop only. Each node in the system is connected to its subsequent node through parallel RF and FSO links. The FSO links and RF links are characterized by Málaga ($\mathcal {M}$) and Nakagami-$m$ composite distributions respectively. For both scenarios, the secrecy outage probability (SOP), strictly positive secrecy capacity (SPSC), intercept probability (IP), and effective secrecy throughput (EST) are obtained as performance metrics. These performance metrics are obtained by considering pointing error, different optical signal detection methods, turbulence effect, and the RF link fading parameter ($m$). Using the results, the security and reliability trade-off analysis is discussed for two scenarios. In addition, the asymptotic results are also obtained to study the system in depth. Further, the secrecy diversity order (SDO) and secrecy coding gain (SCG) are obtained to analyse the system in depth. Finally, Monte-Carlo Simulation is performed to verify the obtained results. The results manifest that the scenario with selection at each hop provides better secrecy performance than the scenario with selection at the last hop.

Kuzushiji, a cursive writing style, had been extensively utilized in Japan for over a thousand years starting from the $8^{th}$ century. In 1900, Kuzushiji was not included in regular school curricula due to the change in the Japanese writing system. Nowadays Japanese natives are unable to read historical books that were written using Kuzushiji language. Therefore, libraries and museums have decided to build digital copies of the documents and books that were written in Kuzushiji language. Due to a limited number of trained experts, researchers have utilized machine and deep learning models to convert historical documents and books into a modern script that can be easily read by human beings. However, the existing deep learning techniques suffer from over-fitting and gradient vanishing problems. To overcome these problems, an efficient deep Kuzushiji characters recognition network (DKNet) is proposed. Initially, to remove noise from the training images, a trilateral joint filter is applied. Contrast limited adaptive histogram equalization (CLAHE) is then applied to enhance the visibility of filtered images. Thereafter, a pre-trained MobileNet is utilized to extract the features of Kuzushiji characters. MobileNet’s final layers are removed, including the fully connected layer and softmax. The flatten layer is then applied to the input. A fully connected classification layer is then used with Rectified linear units (ReLUs) and dropouts. Dropouts are used to generalize the model, thus preventing the over-fitting problem. Finally, the softmax activation function is employed to provide the recognition results. To test the proposed model, actual documents are first segmented by using the proposed Maximally stable extremal regions (MSERs) and convexhull-based segmentation approach. Segmented characters are then recognized using the trained DKNet. Extensive comparative analyses reveal that DKNet achieves better performance than the competitive models in terms of various performance metrics. An efficient Application Programming Interface (API) is also designed for Japanese Kuzushiji ancient heritage character recognition to help the end-users.

In this paper, the physical layer security (PLS) challenges for a decode and forward (DF) based dual hop mixed free space optical (FSO) / radio frequency (RF) communication system in the presence of multiple eavesdroppers has been investigated. In order to improve the PLS of the energy efficient co-operative communication system, diversity combining techniques have been introduced at the destination node. In addition to this, collusion and non collusion eavesdropper scenarios are considered to make the system more practical. The FSO and RF links are assumed to experience independent Málaga ($\mathcal {M}$) and Nakagami-$m$ fading distributions, respectively. Additionally, the FSO link model considers the impact of boresigt and non-zero boresight pointing errors and type of optical demodulation scheme. To investigate the security performance, the lower bound for the secrecy outage probability (SOP) and effective secrecy throughput (EST) have been obtained in terms of Meijer’s G function. Thereafter, asymptotic expressions have been derived for the considered system and the numerical results have been verified by the Monte-Carlo simulation method. Finally, we have analyzed the secrecy energy efficiency (SEE) based on secrecy rate ($R_s$) of the considered system to provide practical insights into the effect of major parameters on the energy efficient secure transmission design.

We investigate the origin of line-shape induced enhancement of stimulated Brillouin scattering (SBS) in narrow linewidth, noise broadened, high-power fiber lasers. A polarization-maintaining seed laser with continuously tunable linewidth (single frequency to >10 GHz), based on white noise modulation was developed for this study. With increasing linewidths, a substantial difference in SBS thresholds was observed depending on the end termination utilized. This observation can be explained by the line-broadened source, having significant power in the Stokes frequency at larger linewidths, seeding the SBS process. Here, SBS threshold for the system terminated with an anti-reflection coated delivery cable is compared with a simple angle cleaved end termination. The influence of end termination on SBS threshold becomes significant with increased linewidths, showing >20% difference in output power between the two cases at ∼10 GHz linewidth. The experiments are complemented by simulations, analyzing relative contributions of Rayleigh scattering and fiber end-facet reflections to SBS. At larger linewidths, due to substantial overlap between laser line-shape and SBS Stokes, with low end-facet reflectivity, Rayleigh is the dominant mechanism, which gives way to end-facet reflections with increasing reflectivity. The Rayleigh contribution is negligible at smaller linewidths, and end-facet reflectivity has a weaker influence than with larger linewidths.

Power scaling of continuous-wave, narrow-linewidth fiber lasers are limited by the onset of stimulated Brillouin scattering (SBS) which manifests at threshold as potentially damaging, peak power, backward propagating pulses. While several SBS mitigation techniques exist, optical linewidth broadening through external phase modulation using white noise is predominantly used for its simplicity. But the broadened line-shape is non-ideal and has a slow roll-off resulting in increased SBS seeding from reflections and hence limited power scaling. With better line-shape control, higher output powers can be achieved. In this work, we propose a simple and effective technique to improve the SBS suppression of these high power, narrow linewidth, noise broadened systems, by the addition of a sinusoidal modulation in cascade. The dual modulation results in an improved line-shape with a relatively flat topped-center and fast roll-off when compared to pure noise modulation at similar root-mean-square (RMS) linewidths. Using a multi-stage fully polarization maintaining, kW-class amplifier, we demonstrate >2.3X enhancement in SBS limited output power when compared to the pure noise modulated case at ∼7.3 GHz and >1.029 kW output power at ∼10.4 GHz RMS linewidth.