Differential-drive mobile robots (DDMRs) are extensively used in logistics applications such as warehouse automation, last-mile delivery, and material handling owing to their simple mechanical structure and high maneuverability. Nevertheless, achieving accurate trajectory tracking remains challenging due to nonholonomic constraints, parameter uncertainties, wheel slip, and external disturbances in dynamic environments. Motivated by the inherent symmetry in the kinematic and dynamic structure of DDMRs, this paper proposes an adaptive control and a fractional-order super-twisting Algorithm (FO-STA)-based control framework for robust trajectory tracking using a dynamic model. The proposed approach adopted a dual-loop control architecture composed of an external kinematic loop and an inner dynamic loop, forming a robust control structure. The kinematic controller ensures convergence of the robot position to the desired reference trajectory, while the dynamic controller compensates for model uncertainties and external disturbances to achieve stable velocity tracking. An adaptive FO-STA mechanism was incorporated to enhance robustness against time-varying dynamics and unknown parameter variations. Both control laws were systematically derived using Lyapunov stability theory, guaranteeing closed-loop convergence and boundedness of tracking errors. Simulation results confirmed the effectiveness of the proposed strategy, demonstrating accurate trajectory tracking and strong robustness under parameter uncertainties and external disturbances, thereby validating its suitability for logistics-oriented mobile robotic systems.

Accurate trajectory tracking in lower-limb exoskeletons is challenged by the nonlinear, time-varying dynamics of human-robot interaction, limited sensor availability, and unknown external disturbances. This study proposes a novel control strategy that combines flatness-based control with two cascaded observers: a high-gain observer to estimate unmeasured joint velocities, and a nonlinear disturbance observer to reconstruct external torque disturbances in real time. These estimates are integrated into the control law to enable robust, state-feedback-based trajectory tracking. The approach is validated through simulation scenarios involving partial state measurements and abrupt external torque perturbations, reflecting realistic rehabilitation conditions. Results confirm that the proposed method significantly enhances tracking accuracy and disturbance rejection capability, demonstrating its strong potential for reliable and adaptive rehabilitation assistance.

Disability is defined as a condition that makes it difficult for a person to perform certain vital activities. In recent years, the integration of the concepts of intelligence in solving various problems for disabled persons has become more frequent. However, controlling an exoskeleton for rehabilitation presents challenges due to their non-linear characteristics and external disturbances caused by the structure itself or the patient wearing the exoskeleton. To remedy these problems, this paper presents a novel adaptive control strategy for upper-limb rehabilitation exoskeletons, addressing the challenges of nonlinear dynamics and external disturbances. The proposed controller integrated a Radial Basis Function Neural Network (RBFNN) with a disturbance observer and employed a high-dimensional integral Lyapunov function to guarantee system stability and trajectory tracking performance. In the control system, the role of the RBFNN was to estimate uncertain signals in the dynamic model, while the disturbance observer tackled external disturbances during trajectory tracking. Artificially created scenarios for Human-Robot interactive experiments and periodically repeated reference trajectory experiments validated the controller’s performance, demonstrating efficient tracking. The proposed controller is found to achieve superior tracking accuracy with Root-Mean-Squared (RMS) errors of 0.022–0.026 rad for all joints, outperforming conventional Proportional-Integral-Derivative (PID) by 73% and Neural-Fuzzy Adaptive Control (NFAC) by 389.47% lower error. These results suggested that the RBFNN adaptive controller, coupled with disturbance compensation, could serve as an effective rehabilitation tool for upper-limb exoskeletons. These results demonstrate the superiority of the proposed method in enhancing rehabilitation accuracy and robustness, offering a promising solution for the control of upper-limb assistive devices. Based on the obtained results and due to their high robustness, the proposed control schemes can be extended to other motor disabilities, including lower limb exoskeletons.

Monitoring and reducing carbon footprints are crucial for achieving sustainability and effectively tackling climate change. Given the urgent global need to address climate change and to participate in Saudi Arabia's Vision 2030, we present DeepCarbonNet (EcoNet), a novel deep learning (DL) framework to monitor, analyze, and reduce carbon emissions. The framework is built using a new Dual Encounter Logarithmic Path Neural Network (DELPNN) architecture, including a novel Spatial Encounter Pathway (SEP), which processes high-resolution satellite images through a Logarithmic Convolutional Encoder (LCE) to extract multi-scale spatial features, and a new Temporal Encounter Pathway (TEP), which processes sequential Internet of Things (IoT) sensor and energy consumption data via a Gated Logarithmic Recurrent Unit (GLRU) and a central Feature Fusion Operator (FFO) that integrates the spatial and temporal features using cross-attention mechanisms and projects them into a logarithmic latent space to capture intricate, non-linear emission dynamics. This approach enables the precise capture of spatial and temporal dependencies within carbon emission data, achieving an outstanding level of accuracy. The simulation experimental results demonstrated that EcoNet attains a high accuracy of 98.7% in estimating carbon footprints after training the model on two public datasets. Furthermore, the model employed a reinforcement learning (RL)-based optimization strategy, enabling a 29.4% reduction in emissions through adaptive mitigation techniques. EcoNet was designed to adapt to changing conditions and promote environmental sustainability continuously. Beyond monitoring, EcoNet achieved a 32.8% improvement in energy efficiency. Additionally, the framework demonstrated robust performance across weather conditions, with 97.0-98.7% accuracy and an accuracy of emission intensities between 94.2–99.1%. These results showed that EcoNet is a solution for artificial intelligence (AI)-driven environmental sustainability, which offers immediate practical value for industrial monitoring, smart city management, logistic services to reduce fuel consumption, and national sustainability programs.

Unmanned aerial vehicles (UAVs) are being increasingly adopted as flexible remote sensing platforms for smart logistics applications, including warehouse inventory, last-mile delivery supervision, traffic flow analysis, port operations, and infrastructure inspection. Despite their advantages, reliable object detection in UAV-based remote sensing imagery remains challenging due to small object sizes, dense object distributions, arbitrary orientations, and complex backgrounds commonly encountered in logistics environments. Although recent YOLO-based detectors have shown promising performance, their effectiveness is often limited in high-resolution aerial scenes and under practical computational constraints imposed by UAV platforms. To address these challenges, this paper proposes DV-YOLO, an enhanced deep learning framework tailored for object detection in UAV-based remote sensing imagery for logistics-oriented applications. The proposed model extends YOLOv9 through a deeper and wider backbone architecture coupled with optimized feature fusion strategies that jointly exploit spatial and semantic representations. A novel cross-path fusion network at deep feature map (CPFNDFM) is introduced to improve the detection of small and densely distributed logistics-related objects such as vehicles, containers, and infrastructure elements. In addition, a lightweight connection aggregation (CA) module, inspired by VoVNet and ShuffleNetV2, is integrated to enhance feature reuse while maintaining computational efficiency suitable for real-time UAV deployment. Furthermore, a challenging benchmark dataset, termed harder vision drone, is constructed by combining and refining samples from VisDrone and DOTA to better reflect real-world UAV remote sensing scenarios in logistics environments. Extensive experimental evaluations conducted on VisDrone 2021, DOTA v2, and the proposed dataset demonstrate that DV-YOLO consistently outperforms state-of-the-art detectors, achieving up to 3.5% improvement in mean average precision (mAP) compared with YOLOv9. These results highlight the potential of the proposed framework to support robust, accurate, and efficient aerial perception for smart logistics and UAV-based remote sensing applications.

Electric vehicles (EVs) are gradually being deployed in the transportation sector. Although they have a high impact on reducing greenhouse gas emissions, their penetration is challenged by their random energy demand and difficult scheduling of their optimal charging. To cope with these problems, this paper presents a novel approach for photovoltaic grid-connected microgrid EV charging station energy demand forecasting. The present study is part of a comprehensive framework involving emerging technologies such as drones and artificial intelligence designed to support the EVs’ charging scheduling task. By using predictive algorithms for solar generation and load demand estimation, this approach aimed at ensuring dynamic and efficient energy flow between the solar energy source, the grid and the electric vehicles. The main contribution of this paper lies in developing an intelligent approach based on deep recurrent neural networks to forecast the energy demand using only its previous records. Therefore, various forecasters based on Long Short-term Memory, Gated Recurrent Unit, and their bi-directional and stacked variants were investigated using a real dataset collected from an EV charging station located at Trieste University (Italy). The developed forecasters have been evaluated and compared according to different metrics, including R, RMSE, MAE, and MAPE. We found that the obtained R values for both PV power generation and energy demand ranged between 97% and 98%. These study findings can be used for reliable and efficient decision-making on the management side of the optimal scheduling of the charging operations.

The transportation and logistics sectors are major contributors to Greenhouse Gase (GHG) emissions. Carbon dioxide (CO₂) from Light-Duty Vehicles (LDVs) is posing serious risks to air quality and public health. Understanding the extent of LDVs’ impact on climate change and human well-being is crucial for informed decision-making and effective mitigation strategies. This study investigates the predictability of CO₂ emissions from LDVs using a comprehensive dataset that includes vehicles from various manufacturers, their CO₂ emission levels, and key influencing factors. Specifically, six Machine Learning (ML) algorithms, ranging from simple linear models to complex non-linear models, were applied under identical conditions to ensure a fair comparison and their performance metrics were calculated. The obtained results showed a significant influence of variables such as engine size on CO₂ emissions. Although the six algorithms have provided accurate forecasts, the Linear Regression (LR) model was found to be sufficient, achieving a Mean Absolute Percentage Error (MAPE) below 0.90% and a Coefficient of Determination (R2) exceeding 99.7%. These findings may contribute to a deeper understanding of LDVs’ role in CO₂ emissions and offer actionable insights for reducing their environmental impact. In fact, vehicle manufacturers can leverage these insights to target key emission-related factors, while policymakers and stakeholders in logistics and transportation can use the models to estimate the CO₂ emissions of new vehicles before their market deployment or to project future emissions from current and expected LDV fleets.