Journal of Fuzzy Systems and Control

Driver Drowsiness Detection and Warning System Using Computer Vision and Neural Networks on Embedded Platforms

Chi-Phat Pham — 2026-06-08

Driver drowsiness is one of the leading causes of traffic accidents worldwide. Traditional monitoring approaches, such as vehicle-based parameter analysis or physiological signal measurement, often require intrusive sensors or deep access to vehicle systems. To overcome these limitations, this paper proposes a real-time driver drowsiness detection and warning system using computer vision combined with a neural network classifier on an embedded platform. Facial landmarks are extracted using the dlib 68-point model, and the Eye Aspect Ratio (EAR) is computed to evaluate eye-closure behavior. A deep neural classifier is trained on eye-state and temporal EAR sequences collected from 25 subjects to classify normal and drowsy conditions. The system is deployed on a Raspberry Pi 3 B+ embedded platform, integrated with an Arduino-based alarm module to deliver audio–visual alerts when drowsiness is detected. Experimental results demonstrate a training accuracy of 98.4% and a testing accuracy of 92.8% with real-time performance of 15–20 FPS under daylight conditions, stable performance in real time, and feasibility for installation in passenger cars, trucks, and buses. The proposed method contributes a low-cost, efficient, and deployable solution for reducing road accidents with a focus on lightweight embedded implementation.

Development of an AI and Webserver-integrated Smart Automated Storage and Retrieval System

Quang-Thien Nguyen — 2026-06-12

In recent years, Automated Storage and Retrieval Systems (AS/RS) and their development have been a notable trend of modern warehouse management by automating the sequential and precise processes of storing, sorting, and retrieving goods. Driven by the convergence of mechatronic systems, Industrial Internet of Things (IIoT), Artificial Intelligence (AI), cloud storage, and edge-based management systems, the potential and practical benefits of AS/RS can be significantly amplified when effectively combined with these trends. In this field, although some works are presented, they often lack specialization for the Vietnamese industrial environment and sustainability. Therefore, this research presents the development of an intelligent AS/RS, incorporating AI-based label processing and webserver-based control to enhance warehouse management efficiency. Experimental evaluations demonstrate that the system achieves high reliability in product classification and storage tasks, providing a scalable solution for modern smart logistics with real-time data synchronization capabilities via a Node-RED web server.

Dual-Stream MobileNetV2 and Light Mixer Fusion for Robust Weather Classification

Muhammad Reza Alfatah — 2026-06-18

This study proposes an image-based weather classification model designed to accurately recognize various sky conditions, with the aim of providing accessible weather information for elderly individuals and users with visual impairments. While existing lightweight models often struggle to effectively capture both fine-grained local textures and global contextual patterns, this study bridges the gap by proposing a hybrid dual-branch architecture. Specifically, the proposed model utilizes an early spatial feature-level fusion dual-branch architecture. The first branch combines MobileNetV2 with a Feature Pyramid Network (FPN) and incorporates selective attention mechanisms to capture multi-scale features. The second branch, referred to as the Mixer branch, improves visual feature representation through patch embedding and feature mixing techniques. Outputs from both branches are integrated using a fusion layer before being processed by a softmax classifier. The dataset includes five weather categories: cloudy, foggy, rainy, sunny, and sunrise, and is preprocessed through normalization, data augmentation, and partitioning into training, validation, and testing sets. Model training is conducted using TensorFlow and Keras with the Adam optimizer over a two-phase training schedule of 60 epochs (20 epochs for head-only pre-training and 40 epochs for whole-model fine-tuning). The experimental evaluation achieves a test accuracy of 0.975 (97.50%), with precision, recall, and F1-score reaching 0.976, 0.975, and 0.975, respectively, reflecting consistent and reliable classification performance. These results indicate that the proposed model has strong potential for integration with text-to-speech systems to improve accessibility of weather information for users with special needs.

Performance Evaluation of Semi-Active Cab Vibration Isolation of a Wheel Loader Using Fractional-Order PID Controller

Chi-Huan Canh — 2026-06-23

Cab vibration in wheel loaders significantly affects operator ride comfort and working performance. Therefore, this paper presents an approach to improving the vibration isolation performance of the cab in a wheel loader system. First, a three-degree-of-freedom dynamic model is established to characterize the vibration behavior. Subsequently, a semi-active cab vibration isolation (SCVI) system is proposed. To generate the semi-active control force, a fractional-order PID (FOPID) controller is developed. Additionally, a conventional PID controller is considered to facilitate a rigorous and comprehensive evaluation of the proposed control strategy. The grey wolf optimization (GWO) algorithm is employed to tune the controller parameters optimally. Finally, the performance of the proposed system is validated through simulations conducted in the MATLAB/Simulink environment. The results indicate that the proposed SCVI system based on FOPID controller reduces the root mean square (RMS) values of seat acceleration (a_zs), cab acceleration (a_zc), and cab vibration isolation mount deflection (z_cf) by 20.76%, 22.33%, and 48.09%, respectively, compared to the passive cab vibration isolation (PCVI) system, demonstrating a significant improvement in operator ride comfort. These findings contribute to the advancement of cab vibration isolation systems for construction machinery.

Adaptive Fuzzy Load Prioritization for Energy Management in Hybrid Wind-Solar Microgrids

Arunava Chatterjee — 2026-06-23

Hybrid renewable microgrids are increasingly becoming popular among renewable energy generation schemes for small-scale power. However, the intermittent nature of both these sources leads to frequent power imbalances between generation and load demand. Conventional control strategies often rely on battery storage for compensation in these cases or employ advanced controls. This paper presents an adaptive Fuzzy logic-based load prioritization strategy for better energy management in such microgrids. The proposed method adjusts the non-critical loads dynamically based on real-time power availability instead of relying on storage only. A fuzzy-based decision technique is used to determine load shedding levels using inputs such as power mismatch, system voltage deviation, and state-of-charge (SOC) of the battery. The proposed control improves system stability and reduces dependency on battery storage. Suitable simulations backed by laboratory-scale experiments demonstrate that the proposed method improves voltage regulation and minimizes load disruption. It significantly reduces voltage deviation by almost 66%, battery current by 43%, and improves power utilization to 97% compared to conventional control systems.

Stabilization of Double Inverted Pendulum using LQR-based Information Fusion Fuzzy Control

Truong-Phuong-Nam Pham — 2026-07-01

Modeling the six-state Double Inverted Pendulum on Cart (DIPC) is highly challenging due to its strong nonlinearities and underactuated dynamics. To address this, the system model in this study is derived using a systematic forward-kinematics-based formulation from robotics theory, previously validated for accuracy in both LQR experiments and ANFIS simulations reported in earlier work. Building on this validated foundation, the present study proposes an Information Fusion Fuzzy Logic Controller (IF-FLC) to overcome the curse of dimensionality commonly encountered when designing fuzzy controllers for high-order systems. The method compresses the six measured state variables into two synthesized linguistic inputs—Synthesized Error (E) and Error Change (EC)—allowing the construction of an efficient 49-rule fuzzy controller without compromising essential system dynamics. Simulations incorporating encoder quantization and realistic measurement constraints show that the proposed IF-FLC provides stable balancing performance and improved robustness compared with the LQR benchmark. The results indicate that information-fusion-based fuzzy design is a promising approach for reducing controller complexity while maintaining high performance, offering a practical pathway for implementing intelligent control strategies on nonlinear and underactuated systems such as the DIPC.