Mr Frej

Mr. Chaouachi Frej — Mechanical Department Supervisor

As our mechanical advisor, Mr. Frej Chaouachi offered structural and functional expertise that laid the foundation of the machine’s physical design:

• Design Feasibility: He guided us through critical decisions related to the structure of the robot, conveyor systems, and liquid-solid separation zones.

• Material Selection & Safety: His advice on using aluminum profiles for protective cages ensured both lightweight durability and operator safety.

• Kinematic Modeling: He reviewed the Delta robot’s kinematics and helped optimize the design for fast, stable motion.

• SolidWorks Modeling: Mr. Chaouachi insisted on clear, clean, and precise CAD modeling—particularly for collision detection, part assembly, and integration with control systems.

His mechanical insights were crucial in transforming our concept into a robust, manufacturable system.

Biography:

Professor Frej Chaouachi is a seasoned academic and project supervisor in the Department of Mechanical Engineering at ENSIT. With deep expertise in mechanical systems design, robotics kinematics, and CAD modeling, he has mentored numerous engineering projects focusing on industrial automation and mechatronics. His approach combines theoretical rigor with practical feasibility, ensuring that student designs are structurally sound, safe, and manufacturable. In this project, his guidance was key to the physical design and structural optimization of the robotic system.

Mr Hassen

Mr. Hassen Seddik was instrumental in power system planning, actuation logic, and automation integration, providing our project with the following strengths:

• Component Selection: He helped us select appropriate electrical components such as servo motors, microcontrollers, and sensors suitable for real-time waste handling.

• Power Management: Advised on current and voltage requirements to ensure the electrical safety and stability of our robot under continuous operation.

• Control Architecture: Provided guidance on the architecture of the control system that governs the interaction between vision, motor action, and robot movement.

• FMEA & Fault Prevention: Encouraged us to apply Failure Mode and Effects Analysis (FMEA) to electrical subsystems to identify risks and optimize reliability.

His electrical expertise enabled the system to be both smart and responsive, with minimal energy waste and fast, accurate actuation.

Biography:

Professor Hassen Seddik is an expert in electrical systems, embedded control, and industrial automation. Serving as a faculty member in the Department of Electrical Engineering at ENSIT, he is known for his methodical approach to power systems, sensor integration, and real-time control architectures. With extensive experience supervising interdisciplinary capstone projects, he ensured the electrical dimension of this system was both stable and optimized for performance, while promoting safe design practices aligned with industry standards.

Ms Bouffayech

Ms. Boufayech’s leadership in the computer science domain drove the intelligence and vision behind the system:

• AI Model Architecture: She guided us in choosing YOLOv11 for object detection, validating our dataset annotation strategy and model configuration.

• Data Annotation & Augmentation: She introduced tools such as Label Studio and recommended advanced preprocessing techniques to improve model generalization.

• Model Evaluation: Ms. Boufayech helped us interpret confusion matrices, recall/precision curves, and F1-scores, ensuring our solution was technically validated and reliable.

• Deployment Strategy: Provided insight on how to transition the AI model from Jupyter-based training environments (like Colab) to real-time edge inference in embedded systems.

Her expertise empowered the system with a reliable and scalable AI core, essential for intelligent automation in real-world conditions.

Biography:

Professor Basma Boufayech is a lecturer and AI researcher within the Computer Science Department at ENSIT. Her fields of interest include computer vision, deep learning, and intelligent systems. She has a strong academic and applied background in designing data-driven models, particularly in detection and classification using convolutional neural networks. In this project, she provided strategic direction on model selection, dataset engineering, and performance evaluation, enabling the integration of an advanced vision-based AI for waste classification.