Globally, potatoes are one of the major crops that significantly contribute to food security; hence, the field of machine learning has opened the gate for many advances in plant disease detection. For real-time agricultural applications, it has been found that real-time data processing is challenging; this is due to the limitations and constraints imposed by hardware platforms. However, such challenges can be handled by deploying simple and optimized AI models serving the need of accurate data classification while taking into consideration hardware resource limitations. Hence, the purpose of this study is to implement a customized and optimized convolutional neural network model for deployment on hardware platforms to classify both potato early blight and potato late blight diseases. Lastly, a thorough comparison between both embedded and PC simulation implementations was conducted for the three models: the implemented CNN model, VGG16, and ResNet50. Raspberry Pi3 was chosen for the embedded implementation in the intermediate stage and NVIDIA Jetson Nano was chosen for the final stage. The suggested model significantly outperformed both the VGG16 and ResNet50 CNNs, as evidenced by the inference time, number of FLOPs, and CPU data usage, with an accuracy of 95% on predicting unseen data.

A novel self-heating technique is proposed to clear snow from photovoltaic panels as a solution to the issue of winter snow accumulation in photovoltaic (PV) power plants. This approach aims to address the shortcomings of existing methods. It reduces PV cell wear, resource loss, and safety risks, without the need for additional devices. A self-heating current is applied to the solar panel to melt the snow covering its surface, which is then allowed to slide off the panel due to gravity. The proposed system consists of a bidirectional DC-DC converter, which removes the snow cover by heating the solar PV modules using electricity from the grid or electric vehicle (EV) batteries. It also charges the EV battery pack and/or supplies the DC bus when no EV is plugged into the charging station. For each mode of operation, a current-controlled system was implemented using a PI controller and a model predictive controller (MPC). The MPC approach achieved a faster rise time, shorter settling time, very low current ripples, and high stability for the proposed system. Specifically, the settling time decreased from 9 ms and 155 ms when using the PI controller at 20 µs and 35 µs with the MPC controller for both the buck and boost modes, respectively

The World Health Organization (WHO) reports that diabetic retinopathy affects one-third of diabetics, regardless of their stage of the disease. Several research efforts are focused on its automated detection and diagnosis. Identifying diabetic retinopathy is crucial due to the damage that occurs to the blood vessels of the eye retina, leading to vision blur or even complete blindness. Thus, an annual checkup is needed for people with diabetes. Moreover, uncontrolled sugar levels for diabetes patients could worsen the current stage of diabetic retinopathy. Consequently, automated detection can greatly contribute to the treatment of disease. This can be carried out through several algorithms, including deep learning models and support vector machines, in addition to transfer learning. This contribution proposes a new approach for diabetic retinopathy automated detection based on convolutional neural network (CNN) models. The proposed model provides both binary and multi-class detection. Both scenarios have shown promising results, where the training accuracies of both the binary classification and the multi-class are 92% and 94%, respectively.