Journal of Robotics and Control (JRC)

An Explainable CNN–LSTM Framework for Monthly Crude Oil Price Forecasting Using WTI Time Series Data

2025-04-20T13:52:35+07:00

Crude oil price forecasting has posed significant challenges due to its volatility and nonlinear dynamics. This study has proposed an explainable CNN–LSTM framework to predict monthly West Texas Intermediate (WTI) crude oil prices. The model has captured both local and sequential patterns without using external inputs or decomposition. Trained over 50 epochs across three data splits, it has been evaluated using RMSE, MAE, MASE, SMAPE, and directional accuracy. A classification accuracy of 92.4% and directional accuracy of up to 87.4% have been achieved. The model has consistently outperformed classical and hybrid baselines, with statistical significance confirmed by the Friedman–Nemenyi test. Saliency-based interpretability has further enhanced transparency, making the framework suitable for real-world energy forecasting.

The Emerging Role of Artificial Intelligence in Identifying Epileptogenic Zone: A Systematic Literature Review

2025-06-08T21:28:34+07:00

Identifying epileptogenic zones (EZs) is a crucial step in the pre-surgical evaluation of drug-resistant epilepsy patients. Conventional methods, including EEG/SEEG visual inspection and neurofunctional imaging, often face challenges in accuracy, reproducibility, and subjectivity. The rapid development of artificial intelligence (AI) technologies in signal processing and neuroscience has enabled their growing use in detecting epileptogenic zones. This systematic review aims to explore recent developments in AI applications for localizing epileptogenic zones, focusing on algorithm types, dataset characteristics, and performance outcomes. A comprehensive literature search was conducted in 2025 across databases such as ScienceDirect, Springer Nature, and IEEE Xplore using relevant keyword combinations. The study selection followed PRISMA guidelines, resulting in 34 scientific articles published between 2020 and 2024. Extracted data included AI methods, algorithm types, dataset modalities, and performance metrics (accuracy, AUC, sensitivity, and F1-score). Results showed that deep learning was the most used approach (44%), followed by machine learning (35%), multi-methods (18%), and knowledge-based systems (3%). CNN and ANN were the most commonly applied algorithms, particularly in scalp EEG and SEEG-based studies. Datasets ranged from public sources (Bonn, CHB-MIT) to high-resolution clinical SEEG recordings. Multimodal and hybrid models demonstrated superior performance, with several studies achieving accuracy rates above 98%. This review confirms that AI (especially deep learning with SEEG and multimodal integration) has strong potential to improve the precision, efficiency, and scalability of EZ detection. To facilitate clinical adoption, future research should focus on standardizing data pipelines, validating AI models in real-world settings, and developing explainable, ethically responsible AI systems.

Dynamic Clustering of Multi-Mobile Robot System using Gaussian Mixture Model

2025-05-31T22:44:24+07:00

Managing large fleets of mobile robots poses significant challenges to system coordination and workload. An effective grouping strategy is crucial for enhancing operational performance and scalability. This paper introduces a two-stage dynamic clustering method (DCM), a novel framework for organizing robots into manageable groups. The methodology utilizes a Gaussian Mixture Model and the Expectation-Maximization algorithm to cluster robots based on their path intersection points. A unique "cost" parameter, formulated a least squares objective function, is proposed to guide the selection of near-optimal, workload-balanced configurations. The results from extensive simulations demonstrated the framework's effectiveness. On a single dataset, DCM exhibited exceptional reliability, maintaining a stable objective function value even as the number of robots per cluster fluctuated across runs. A sensitivity analysis over multiple unique datasets confirmed the model's adaptive strength, showing its ability to re-configure clusters. This adaptability was highlighted by the mean objective function value varying across different scenarios. Further analysis involving reduced robot populations and obstacle-filled environments validated DCM's generalizability and environment-independent nature. The robot distribution mechanism was consistently equitable and balanced. Statistical validation, including bootstrapping resamples, confirmed the stability and reliability of the performance estimates. The method also steadily maintained a high level of performance by adapting to internal variations. Moreover, every robot was successfully assigned to all clusters across all trials. The research concludes that DCM is a robust, adaptive, and environment-independent framework. It successfully balances performance stability with the flexibility to respond to new operational conditions, proving it is an effective solution for multi-robot coordination.

Computer Vision for Food Nutrition Assessment: A Bibliometric Analysis and Technical Review

2025-06-28T13:54:52+07:00

This study examines the latest trends, challenges, and advances in food image segmentation and computer vision-based nutritional analysis. Traditional nutritional assessment methods such as food diaries and questionnaires are limited by their reliance on participant recall and manual processing, which reduces their accuracy and efficiency. As an alternative, advances in machine learning and deep learning have shown potential in automating food identification and estimating nutrient content, such as calories, protein, carbohydrates, and fat. This study was conducted through bibliometric analysis and technical review of publications from the Scopus database, using a structured search strategy and applying inclusion and exclusion criteria. Articles were selected based on topic relevance, use of machine learning or deep learning methods, publication in English, and publication between 2020 and 2024. The review identified key research trends, key contributors, popular methods such as CNN and YOLO, and the most frequently reported limitations, including lack of dataset diversity, inaccuracy in food volume estimation, and the need for real-time integrated systems. These limitations were analyzed based on the methodology and findings of the reviewed studies. This review is expected to be a comprehensive reference for researchers and practitioners in developing food image segmentation technology for more accurate and applicable nutritional assessment.

Predicting Occupational Heat Stress in Critical Sectors: A Sector-Based Systematic Review of Wearable Sensing, IoT Platforms, and Machine Learning Models

2025-06-17T05:51:47+07:00

Occupational heat stress is a growing threat to the health and productivity of workers exposed to extreme environmental conditions. This issue is particularly acute in sectors such as construction, mining, agriculture, and heavy industry, where high heat exposure and physical workload are constant. This systematic review analyzes 96 scientific articles published in recent years, aiming to identify emerging technological systems focused on the prediction, monitoring, and mitigation of occupational heat stress. The main contribution of this study lies in the cross-sectoral categorization of recent solutions, providing a comparative framework that highlights knowledge gaps, methodological limitations, and opportunities for innovation. Following PRISMA guidelines, data were extracted on sensor type, predictive models, validation environments, and the sector of application. Technologies were classified into five main categories: wearable sensors, IoT-based monitoring platforms, hybrid thermal indices, predictive models based on environmental and physiological inputs, and decision-support tools. The results reveal a strong presence of wearable systems. Adoption is further constrained by socio-technical barriers such as worker compliance, PPE burden, costs, data privacy, and interoperability gaps. However, only a small fraction of studies conducted in-field validation under real thermal stress conditions, and even fewer included longitudinal ergonomic trials, limiting generalizability, with additional concerns about heterogeneous outcome measures and inconsistent definitions of heat stress across studies. A sectoral imbalance is also observed, with construction and industrial environments receiving more research attention than mining, agriculture, and indoor workplaces. In conclusion, we propose a practical roadmap for the adoption of standardized data schemas and protocols, field trials across complete work cycles, privacy-preserving analytics (federated learning), and integration of ergonomic and organizational controls. In highly humid or high radiation settings, complementing or replacing WBGT with hybrid indices (UTCI) can improve risk estimation and enable more actionable work rest and hydration alerts.

Sensor Fusion and Predictive Control for Adaptive Vehicle Headlamp Alignment: A Comparative Analysis

2025-04-29T13:57:29+07:00

Nighttime driving safety is often compromised by the inability of conventional adaptive headlamp systems to account for lateral slip and rapidly changing road conditions, leading to misalignment and reduced visibility during aggressive maneuvers. Most existing approaches rely solely on steering angle, which limits adaptability under dynamic slip scenarios. This study presents the development and comparative evaluation of a Fused Controller that uniquely integrates sensor fusion, adaptive gain scheduling, and multi-step predictive optimization for robust adaptive headlamp alignment. Five control architectures- Filtered Proportional Controller (FPC), Raw State MPC (RS-MPC), Extended MPC (E-MPC), Feedforward-Enhanced MPC (FF-MPC), and the proposed Fused Controller- were systematically evaluated on a 2 km synthetic road with ten challenging segments. Compared to the E-MPC baseline, the Fused Controller achieved a 42.5% reduction in root mean square error (RMSE) in long S-curves and a 30.6% improvement in sharp turns, with a settling time of 0.6 s (versus 1.8 s for FPC) and a jitter index of 9.93°/s. Frequency-domain analysis confirmed a 1.2 Hz bandwidth with actuator-compatible roll-off, and stability analysis validated robustness under noise and disturbances. Statistical analysis across 20 independent simulation runs per controller showed these improvements are highly significant (p < 0.001, large Cohen’s d), confirming the practical superiority of the Fused Controller. These results indicate enhanced driver visibility and reduced nighttime collision risk, while the controller’s computational efficiency and adaptive gains support scalability and real-world deployment. This work provides a rigorous and practical framework for next-generation adaptive lighting systems.

Robotics-Driven Biometric Authentication for Secure and Intelligent Vehicle Access

2025-03-02T11:38:19+07:00

As modern vehicles increasingly adopt intelligent systems, the need for robust and secure access control mechanisms has become paramount. This study presents a dual modal biometric authentication framework integrating fingerprint and iris recognition to enhance vehicular security and user convenience. The system leverages strategically positioned sensors—capacitive fingerprint scanners embedded within door handles and high-resolution iris scanners mounted near the driver’s entry point—coupled with a central microcontroller for real-time processing. Lightweight image processing and matching algorithms are implemented to ensure fast and accurate authentication under varied environmental conditions. The proposed system was validated on a prototype vehicle model using biometric data from over 50 users, demonstrating high accuracy, low false acceptance/rejection rates, and resilience to spoofing attacks. In addition to technical implementation, the study addresses practical challenges including sensor placement, processing constraints, data privacy, and system usability. The findings support the feasibility of integrating multimodal biometric authentication in vehicles, offering a secure, user-friendly alternative to conventional key- based systems. This research is supported by Indian patent file number 202441019532.

The Impact of Photovoltaic Systems on the Performance of Induction Motor in Agricultural Irrigation Applications

2025-06-17T05:53:12+07:00

Water is a vital resource in the agricultural sector, as most farmland relies on tubewells for irrigation. Solar photovoltaic pumping systems (SPVPS) have emerged as a promising solution for sustainable agricultural irrigation, providing clean and efficient alternatives to traditional energy sources. However, induction motors used in (SPVPS) suffer from problems including voltage instability, decreased efficiency, overheating, and mechanical stress due to the varying nature of PV electricity. This paper focuses on the analysis of an irrigation system by MATLAB/Simulink simulation environment to analyze the performance of (SPVWPS), which consists of a 22 kW three-phase induction motor connected to a photovoltaic system under the climatic operating conditions of Mosul, Iraq. The results showed that operating the water pump using the solar system led to a decrease in the motor torque by 9% and the motor efficiency decreased by 34.3% compared to when supplied with electrical power from the grid. The results showed that the system achieved its best performance at a peak of the solar irradiance of 860 W/m² and a temperature of 24.6°C, with the induction motor speed reaching 1,317 rpm and a maximum efficiency of 48.4%. The total harmonic distortion (THD) in the rotor current peaked at 184.37% at 24°C before decreasing to approximately 45.3% at higher temperatures. The increased THD is due to a combination of inverter stress, poor waveform quality under thermal load, and high-frequency disturbances caused by variable environmental conditions. It’s a typical challenge in solar-powered motor drives, especially in off-grid or remote agricultural systems. These results effectively contribute to supporting sustainable agricultural practices by ensuring the continuity and efficiency of motor operation.

Boosting Energy for Building-Integrated Photovoltaic Cells using Novel Boost Converter with Voltage Multiplier Cell and ANN-MPPT

2025-06-17T06:04:15+07:00

This study investigates optimizing photovoltaic (PV) energy delivery to building lighting loads by proposing a novel boost converter with a voltage multiplier stage (VMS) and an intelligent maximum power point tracking (MPPT) system. The research contribution is the design and comparative analysis of this advanced converter topology against a traditional boost converter to demonstrate enhanced performance under diverse operating conditions. The methodology involves simulating the PV system under four distinct scenarios including variations in load resistance, desired output voltage, and dynamic solar irradiance. The performance of three MPPT algorithms namely artificial neural network (ANN), particle swarm optimization (PSO), and perturb and observe (P&O), was evaluated to identify the most effective control strategy. The results by using MATLAB/Simulink show that the proposed boost VMS converter consistently outperforms the traditional boost converter by exhibiting improved power extraction and enhanced stability in output voltage and current. For example in a scenario with a 50 V output and 1000 W/m² irradiance the boost VMS converter achieved a more stable output power of approximately (961.52W) compared to (941.543W) from the traditional converter. Furthermore the ANN-based MPPT demonstrated superior stability and power tracking accuracy especially under dynamic irradiance conditions, where it maintained a stable output while PSO and P&O experienced significant power drops. Integrating the boost VMS converter with an ANN-based MPPT provides a superior, robust solution for optimizing PV energy utilization in building lighting applications, ensuring efficient and stable power delivery under fluctuating environmental and load conditions.

Lightweight Lattice-Based Multi-Domain Authentication Protocol with Real-Time Revocation and Aggregated Verification for Vehicular Communication

2025-04-23T17:13:28+07:00

Vehicle-centric vehicular communication systems need secure, scalable, and low-delay authentication schemes to guarantee on-the-fly trust among vehicles, roadside units (RSUs), and cloud services. The research contribution is a authentication with which the quantum entities of appropriate domains exchange the quantum messages to achieve the quantum resistance and the vehicular authentication among the multi-domains. We design an efficient lattice-based authentication scheme spanning Ring-LWE for post-quantum key generation, ring signatures for anonymity, and Merkle tree structures for space-efficient public key management. Merklix trees can be anchored to combat decentralized and globally verifiable revocation using a consortium blockchain. To cope with high-density traffic, we devise an aggregated verification approach to minimize the computational and communication cost. The scheme functions in four stages-initialization, registration, mutual authentication and revocation together with pushing the real-time alert based on the compromised key. The security is reduced to the Random Oarch Model (ROM), with hardness assumptions defined over the hard lattice problems such as CBi-ISIS and Ring-LWE. Our simulation results on the realistic vehiculargrade devices demonstrate that our protocol can readily achieve sub-25 ms authentication latency, small-size signature (1.3 KB) and convenient Merkle proof processing procedures, which outperforms the state-of-the-art lattice-based schemes. These findings indicate the feasibility of the system for real-time V2X services. It is validated to provide scalable, privacy-preserving authentication for packed vehicular networks. Next, we plan to investigate the adaptive trust scoring and dynamic batch verification for the mobility.

Cooperative Lane Keeping Assist: Design and Evaluation of a V2V Lane Perception Sharing Approach

2025-05-02T01:21:18+07:00

Even autonomous vehicles are becoming very advanced. Adverse weather conditions, unclear lane markings, and unexpected obstacles can still pose challenges, especially to lane keeping assist systems. The performance of these systems varies between vehicles depending on sensor quality, environmental conditions, and data processing algorithms. A focused solution to improve lane keeping capability is vehicle-to-vehicle (V2V) communication. V2V enables vehicles to share real-time information on speed, position, direction, etc. In this paper, V2V is used specifically to share lane marking data from a front vehicle to a following vehicle. These data are fused with local perception using a confidence-weighted averaging method, where each lane position input is assigned a confidence score. When local perception degrades, such as during poor weather, this approach improves lane keeping by relying on the more reliable lane marking positions of the front vehicle. We validate our V2Venhanced LKA system using MATLAB/Simulink simulations with one front vehicle. Results show up to a 92.75% reduction in mean error compared to standard LKA and smoother steering. Since the system shares only lane marking positions for lane keeping purposes, the communication load remains low. However, attention must still be given to cybersecurity aspects, as even limited data exchange via V2V is vulnerable to threats such as spoofing or tampering, which could compromise the safety of the lane keeping function.

Adaptive Trajectory Control for Quadcopter using Extended Kalman Filter-Based Self-Tuning PID under Gaussian Disturbances

2025-07-22T10:31:05+07:00

Quadcopters are known for their maneuverability, but their stability is often challenged in changing environments. The PID parameters are adjusted manually so it is less adaptive. This research introduces the combination of Self-Tuning Proportional-Integral-Derivative (PID) and Extended Kalman Filter (EKF). A PID controller adjusts parameters based on errors and state estimates obtained from the EKF in real time. The disturbances used are Gaussian random disturbances on the system and sensors, simulated a normal distribution using the Simulink model. The basic PID parameters are determined through numerical simulations, then adaptively calibrated with a multiplier function based on estimation and error. The contributions of this study are: (1) developing an EKF-based SelfTuning PID control for a quadcopter system; (2) demonstrating an adaptive response to disturbances through simulations; and (3) presenting an efficient tuning strategy suitable for resourcelimited systems. From simulation, z-axis overshoot is successfully decreased from 7.37% to only 2.54%, while the steady-state error remains low under system disturbances. Computational efficiency is achieved because 12 state variables are controlled using a single set of global PID parameters, and the tuning process takes place in real time without relying on complex AI-based optimization methods. The proposed control approach is able to maintain trajectory tracking accuracy in three-dimensional space adaptively and with efficient resource usage. These results demonstrate that the EKF-PID method is effective for UAV control in dynamic and disruptive environments.

Enhancing Network Lifetime and Data Integrity in WSNs via Optimized Mobile Robot Trajectories

2025-04-20T13:52:24+07:00

Recent research has shown that utilizing mobile robot data collection from sensor nodes is one of the most critical schemes to prolong the network lifetime in wireless sensor networks (WSNs). By overcoming some limitations of traditional methods where sensing data is sent to a static data collection node through multiple routing paths, the mobile data collection robot-based approaches can completely avoid "hotspot" problem, energy-holes issues thereby balancing node energy consumption in the network. Consequently, many ideas and publications on improving network lifetime in WSNs by utilizing mobile data collection robot(s) have been proposed. However, there is little research that has studied the impact of mobile robot trajectory types on network lifetime improvement. Therefore, it becomes very interesting to investigate data collection process of mobile robots in wireless sensor network. In this paper, we proposed a geometric solution to find optimal trajectories of utilized mobile robots (MRs). Our proposed solution consists of four main stages. In the first stage, the number of cluster head nodes is estimated based on the network size and the density of sensor nodes in the WSN. The second stage involves estimating the spatial region that each mobile robot must cover to collect sensed data from all assigned sensor nodes. In the third stage, an optimal trajectory for each mobile robot is determined. In the fourth and final stage, the Network Control Center (NCC) proceeds to assign optimized trajectories to the remaining mobile robots until all cluster head nodes in the network have been visited. The proposed optimal trajectory for the mobile robot is designed not only to ensure timely collection of all sensed data in the field, but also to minimize the energy consumption of sensor nodes, thereby improving the overall network lifetime. A large number of numerical tests were carried out to evaluate the performance of our proposed algorithm. The simulation results demonstrate that our proposed algorithm achieves a 5.4% improvement in network lifetime compared to other traditional algorithms. Nevertheless, the network lifetime improvement remains dependent on several assumptions made in this study. To address this limitation, the discussion section of the paper outlines potential directions for future work aimed at enhancing the practical applicability of the proposed solution.

Reinforcement Learning for Multi-Task Manipulation in Robotic Arm Systems Operating in Dynamic Environments

2025-07-12T13:55:51+07:00

The development of integrating Reinforcement Learning (RL) in robots seems to provide solutions to a variety of complex manipulation tasks in uncertain dynamic environments. The limitation of the research is that the given research permits a robotic arm to learn and perform several manipulation tasks in an autonomously-observed manner, using a model-free RL approach. The key improvement of the current work is an ability to train an agent to perform various actions in a shared space that is needed to perform very different manipulation actions. The method is implemented with the help of a three-dimensional simulator that is done by using a robotic arm, items in the workspace (table, objects), and the time-varying location of targets. The robotic system undergoes training in six different manipulation actions including Reach, Push, Slide, Pick and Place, Stack and Flip. With reward shaping based on the tasks, RL architecture learns to execute each task effectively in working with the environment. The success rates in each of the manipulation tasks during experiment time have demonstrated successful completion of the tasks as opposed to before training, displaying their adaptability and accuracy. Also, this framework has generalization ability as it changes its object positions and the dynamics used. These results help justify the possibility of reinforcement learning as a tool to train robots on flexible, goal directed manipulation tasks and so avoid manual programming. Future work may extend this approach to real-world robotic platforms with sensory feedback integration.

Deep Q-Network-Based Path Planning in a Simulated Warehouse Environment with SLAM Map Integration and Dynamic Obstacles

2025-06-28T10:49:35+07:00

With the rise of e-Commerce and the evolution of robotic technologies, the focus on autonomous navigation within warehouse environments has increased. This study presents a simulation-based framework for path planning using Deep Q- Networks (DQN) in a warehouse environment modeled with moving obstacles. The proposed solution integrates a prebuilt map of the environment generated using Simultaneous Localization and Mapping (SLAM), which provides prior spatial knowledge of static obstacles. The reinforcement learning model is formulated with a state space derived from grayscale images that combine the static map generated by SLAM and dynamic obstacles in real time. The action space consists of four discrete movements for the agent. A reward shaping strategy includes a distance-based reward and penalty for collisions to encourage goal-reaching and discourage collisions. An epsilon-greedy policy with exponential decay is used to balance exploration and exploitation. This system was implemented in the Robot Operating System (ROS) and Gazebo simulation environment. The agent was trained over 1000 episodes and metrics such as the number of actions executed to reach the goal and the cumulative reward per episode were analyzed to evaluate the convergence of the proposed solution. The results across two goal locations show that incorporating the SLAM map enhances learning stability, with the agent reaching a goal approximately 150 times, nearly double the success rate compared to the baseline without map information, which achieved only 80 successful episodes over the same number of episodes. This indicates faster convergence and reduced exploration overhead due to improved spatial awareness.