We received many outstanding graduation projects from different regions, each demonstrating strong technical depth and meaningful societal relevance.
Winners' selection was not an easy task !
We are glad to celebrate the winners of the 2024–2025 edition of the Challenges, whose projects primarily prioritize local relevance, innovation, and societal impact. The winning projects stood out for their proposed solutions to tackle local challenges, especially those developed within an industrial hosting environment.
Congratulations to the winners and a sincere appreciation to all participants.
Thanks for every submission that contributed valuable ideas and meaningful perspectives.
For his graduation project:
"Application of Convolutional Neural Networks and Vision Transformers in Cancer Grading in Pathology Images"
Project Summary
This project introduces an automated deep learning system to grade colon cancer from medical images, addressing the inconsistencies and heavy workload of manual doctor evaluations. To improve accuracy, Dosbol developed a novel neural network architecture called "Post-Convolutional Parallel CBAM." Unlike standard methods that can disrupt data flow, this new approach processes different image features simultaneously. When tested on a standard medical dataset, this parallel method significantly outperformed existing models. Notably, for the VGG16 model, the accuracy (F1-score) jumped from 0.658 to 0.810. Furthermore, this innovative approach is highly efficient, requiring roughly 300,000 fewer computational parameters on models like GoogLeNet. Ultimately, this system provides a faster, more reliable, and lightweight tool to assist medical professionals in cancer diagnosis.
For her graduation project:
"Design and Implementation of a Medical Diagnosis System Based on Deep Learning and Signal Processing for ECG Analysis"
Project Summary
This project presents the design and implementation of a portable, AI-powered medical diagnostic kit for real-time ECG signal acquisition, processing, and analysis. Imen designed a system composed of a transmitter and a receiver, combining low-cost embedded hardware with advanced deep learning techniques. The transmitter uses a custom PCB based on an ESP32 microcontroller to acquire biomedical signals such as ECG and PPG. The receiver, built on an Orange Pi, performs signal preprocessing, AI-based heartbeat abnormality detection, and visualization through a web dashboard. This end-to-end architecture connects real-time data acquisition with intelligent analysis in a compact system. The prototype enables real-time ECG monitoring and automated diagnosis support. This project demonstrates the feasibility of affordable, intelligent, and portable cardiac diagnostic tools.
For her graduation project:
"Optimization of Energy Consumption in Data Centers through Photovoltaic Integration and Green Cooling Solutions."
Project Summary
This project focuses on optimizing the energy consumption of a large-scale data center through the integration of renewable energy and advanced cooling solutions. A 1.2 MW photovoltaic system was designed and simulated to ensure both technical performance and economic viability. In parallel, an innovative green cooling strategy was developed, combining free cooling techniques with server heat recovery to enhance overall efficiency. The project not only reduces operational costs but also significantly lowers the carbon footprint of the data center. By integrating sustainable energy and intelligent thermal management, this work contributes to improving energy efficiency while aligning data center operations with environmental and sustainability goals.
