In this study, an automated multimodal system for detecting, classifying, and dating fruit was developed using a two-stage YOLOv11 pipeline. In the first stage, the YOLOv11 detection model locates individual date fruits in real time by drawing bounding boxes around them. These bounding boxes are subsequently passed to a YOLOv11 classification model, which analyzes cropped images and assigns class labels. An additional counting module automatically tallies the detected fruits, offering a near-instantaneous estimation of quantity. The experimental results suggest high precision and recall for detection, high classification accuracy (across 15 classes), and near-perfect counting in real time. This paper presents a multi-stage pipeline for date fruit detection, classification, and automated counting, employing YOLOv11-based models to achieve high accuracy while maintaining real-time throughput. The results demonstrated that the detection precision exceeded 90%, the classification accuracy approached 92%, and the counting module correlated closely with the manual tallies. These findings confirm the potential of reducing manual labour and enhancing operational efficiency in post-harvesting processes. Future studies will include dataset expansion, user-centric interfaces, and integration with harvesting robotics.

Accurate vehicle localization is essential for safety-critical vehicular ad hoc networks (VANETs), including emergency response, navigation, traffic monitoring, and cooperative driving. However, conventional GPS/GNSS positioning systems have often shown degradation in tunnels, dense urban corridors, and non-line-of-sight (NLOS) environments, where satellite visibility and signal reliability are limited. This paper proposes a calibrated Link Quality Indicator (LQI)-based closed-form localization framework for partially connected Roadside Unit (RSU)-assisted VANETs. The proposed framework first calibrates the LQI-to-range relationship using numerical regression parameters and then converts accepted LQI observations into distance estimates. The distances are processed through a variance-aware weighted least squares (WLS) estimator, after which scalar consistency refinement is applied to improve the geometric consistency of the final position estimate. Partial connectivity is explicitly modeled using communication radius, LQI-threshold, and packet-reception constraints, rather than assuming that all vehicles and anchors are fully connected. The proposed method is evaluated against least squares (LS), WLS, RSSI-WLS, Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), Particle Filter (PF), cooperative localization, hybrid GNSS/INS/RSS fusion, and machine-learning-based localization baselines. The evaluation includes Monte Carlo simulations under shadowing, fading, packet loss, anchor-geometry, and NLOS conditions, together with confidence interval, statistical-significance, runtime, ablation, and sensitivity analyses. In addition, trace-driven validation is performed using the NGSIM US-101 real-world vehicle trajectory dataset. The NGSIM dataset provides real vehicle mobility traces, while LQI observations are generated using the calibrated LQI-distance model because the dataset does not contain physical LQI measurements. In the 100-vehicle case, throughput improves from 0.614 to 1.169 successful localizations/s compared with LS, corresponding to a 90.39% gain and from 0.591 to 1.169 successful localizations/s compared with WLS, corresponding to a 97.80% gain. In the NGSIM trace-driven experiment, the proposed method achieves RMSE values of 2.39, 2.55, and 3.14 m for 10, 50, and 100 vehicles, respectively. These results indicate that calibrated LQI-based localization provides a low-cost, infrastructure-compatible, and computationally efficient positioning framework for ITS and safety-critical VANET applications.

Most predictive maintenance studies have emphasized accuracy but provide very little focus on Interpretability or deployment readiness. This study improves on prior methods by developing a small yet robust system that can predict when turbofan engines will fail. It uses the NASA CMAPSS dataset, which has over 200,000 engine cycles from 260 engines. The process begins with systematic preprocessing, which includes imputation, outlier removal, scaling, and labelling of the remaining useful life. Dimensionality is reduced using a hybrid selection method that combines variance filtering, recursive elimination, and gradient-boosted importance scores, yielding a stable set of 10 informative sensors. To mitigate class imbalance, minority cases are oversampled, and class-weighted losses are applied during training. Benchmarking is carried out with logistic regression, gradient boosting, and a recurrent design that integrates gated recurrent units with long short-term memory networks. The Long Short-Term Memory–Gated Recurrent Unit (LSTM–GRU) hybrid achieved the strongest performance with an F1 score of 0.92, precision of 0.93, recall of 0.91, Receiver Operating Characteristic–Area Under the Curve (ROC-AUC) of 0.97, and minority recall of 0.75. Interpretability testing using permutation importance and Shapley values indicates that sensors 13, 15, and 11 are the most important indicators of engine wear. The proposed system combines imbalance handling, feature reduction, and Interpretability into a practical design suitable for real industrial settings.