Reducing product damage, preserving quality, and enhancing efficiency from harvest to consumption are crucial for sustainable agriculture. The integration of advanced information and communication technologies into agricultural practices plays a vital role in meeting these goals. This study introduces an autonomous transport vehicle designed for the efficient logistics of fruit transportation in agricultural settings . The vehicle's software framework is constructed on the Robot Operating System (ROS) and incorporates an enhanced hybrid navigation system that merges the Extended Kalman Filter (EKF) with Simultaneous Localization and Mapping (SLAM) for precise localization. The A* algorithm facilitates global path planning, whereas the Dynamic Window Approach (DWA) guarantees real‐time obstacle avoidance. Essential hardware components comprise high‐resolution LIDAR for environmental mapping, an Inertial Measurement Unit (IMU) for motion estimation, and wheel encoders for odometry. The performance evaluation was executed across five distinct terrain types: concrete, fine‐tilled soil, coarse‐tilled soil, asphalt, and grass. The vehicle attained optimal path‐following precision on concrete, exhibiting a deviation of 5.39 cm at a speed of 0.3 m/s with a 200 kg payload, whereas tracking errors escalated on uneven terrains like grass and coarse‐tilled soil. Maneuverability assessments verified a turning radius of 60.0 cm for 90° turns and 125.0 cm for 180° turns, ensuring suitability in restricted agricultural environments. Finite element analysis (FEA) evaluated structural durability under diverse loads (2000–4000 N), indicating a minimum safety factor of 1.23, thereby affirming structural stability under static conditions. This study demonstrates the potential of autonomous transport vehicles to revolutionize agricultural logistics by reducing labor dependency, improving operational efficiency, and supporting sustainable farming.

In this study, an image processing algorithm and classification unit were developed to classify the fruits according to their size and color characteristics. For this purpose, a total of 300 fruits (50 fruit samples from each of the Starkrimson Delicious and Golden Delicious apple varieties, Washington Navel and Valencia Midknight orange varieties, Ekmek and Eşme quince varieties) were used in the experiments. The size and color values measured with a caliper and a spectrophotometer were entered in the developed image processing algorithm to determine the success rates of classifying the fruits. The integration of image processing algorithm with the classification unit classified 88%, 100%, 96%, 82%, 86%, respectively. On the other hand, the size and color values read in fruits with the image processing algorithm were evaluated using predictive techniques used in data mining. For this purpose, K Nearest Neighbor (KNN), Decision Tree (DT), Naive Bayes classification and Multilayer Perceptron Neural Network (MLP) algorithms were used. Algorithms were run with 10-fold cross validation method. In the training of artificial classifiers, the success was 93.6% for KNN, 90.3% for DT, 88.3% for Naive Bayes, 92.6% for MLP and 94.3% for RF.

Tomato is one of the most important vegetables in the world. Presence of diseases and pests in the growing area significantly affect the choice of variety in tomato. Early-stage diagnosis plays an important role in determining whether the tomato is subject to effective and economical chemical, mechanical and biological controls, and internal and external quarantine. In this study, deep learning was used to diagnose some diseases in tomatoes. For this purpose, a novel deep CNN-based approach and some Keras models including DenseNet201, InceptionResNetV2, MobileNet, VGG16 architectures were used. Early, middle, and late stages of 18.456 images of Bacterial Spot, Early Blight, Leaf Mold, Septoria Leaf Spot, Target Spot, Mosaic Virus, Yellow Leaf Curl Virus and healthy leaves were examined. The experimental results showed that the custom CNN model produced 99.82% training accuracy. We recommend this model in terms of monitoring and diagnosing of tomato diseases. The results obtained with this study can be used in robotic spraying and harvesting operations.

Image analysis techniques are developing as applicable to the approaches of quantitative analysis, which is aimed to determine cultivar grains. Additionally, corn (Zea mays) grain processing companies evaluate the quality of kernels to determine the price of these cultivars. Because of this reason, in the study, a computer image analysis technique was applied on three corn cultivars. These were Zea mays L. indentata, Zea mays L. saccharata and a hybrid corn (Yellow sweet corn). These cultivars are commercially important as dry grains in Turkey. In the study, the grain color values were tested in the cultivars from Turkey’s collection. One hundred samples were used for each corn cultivar, and 300 corn grains in total were used for evaluations. Each of nine color parameters (Rmin, Rmean, Rmax, Gmin, Gmean, Gmax, Bmin, Bmean, Bmax) which were obtained from original RGB color channels with maximum and minimum values was evaluated from the digital images of three different corn cultivar grains. The values were analyzed with the help of the Multilayer Perceptron (MLP), Decision Tree (DT), Gradient Boost Decision Tree (GBDT) and Random Forest (RF) algorithms by using the Knime Analytics Platform. The majority voting method was applied to MLP and DT for prediction fusion. All algorithms were run with a 10-fold cross-validation method. The success of prediction accuracy was found as 99% for RF and GBDT, 97.66% for MLP, 96.66% DT and 97.40% for Majority Voting (MAVL). The MAVL method increased the accuracy of DT while decreasing the accuracy of MLP partly for the fusion of MLP and DT.

Classification of hazelnuts is one of the values adding processes that increase the marketability and profitability of its production. While traditional classification methods are used commonly, machine learning and deep learning can be implemented to enhance the hazelnut classification processes. This paper presents the results of a comparative study of machine learning frameworks to classify hazelnut (Corylus avellana L.) cultivars (‘Sivri’, ‘Kara’, ‘Tombul’) using DL4J and ensemble learning algorithms. For each cultivar, 50 samples were used for evaluations. Maximum length, width, compression strength, and weight of hazelnuts were measured using a caliper and a force transducer. Gradient boosting machine (Boosting), random forest (Bagging), and DL4J feedforward (Deep Learning) algorithms were applied in traditional machine learning algorithms. The data set was partitioned into a 10-fold-cross validation method. The classifier performance criteria of accuracy (%), error percentage (%), F-Measure, Cohen’s Kappa, recall, precision, true positive (TP), false positive (FP), true negative (TN), false negative (FN) values are provided in the results section. The results showed classification accuracies of 94% for Gradient Boosting, 100% for Random Forest, and 94% for DL4J Feedforward algorithms.