Mariem Ben Slimene

Conducted in-depth analysis of Wallbox EV Chargers, exploring functionalities, features, and performance metrics. • Tested and debugged hardware and software components to identify issues and ensure reliable operation. • Evaluated potential real-world applications of Wallbox functionalities, proposing practical integration strategies. • Coordinated findings with engineering and operations teams, supporting informed project decisions. • Maintained detailed technical records and reports, documenting tests, issues, and recommendations for management review. Key words: PV-optimized charging integration, OCPP, Modbus RTU/TCP, Dynamic and static load/charge management, Current limitation, Over/under voltage control, Debugging, Technical documentation.

Project : Sensor Calibration leveraging by AI. • Accelerated the calibration of Infineon’s TLE4971 automotive current sensors by eliminating 33% of testing, reduce costs and production time. • Analyzed initial approaches available already. • Translate existing KNIME Workflows, into Python. • Developed XGBoost models to predictcalibration. • Implemented feature selection from data points, and hyperparameters tuning. • Compared different results and validate results against production standards.

Project : Soil moisture prediction using LSTM model. • Address Tunisia’s water scarcity by optimizing irrigation through AI-driven soil moisture prediction, reducing agricultural water waste. • Built an LSTM model to predict soil moisture at 5depths using digitized sensor data and NASA weather data. • Trained and evaluated the model’s performance using RMSE, focusing on capturing long term dependencies in soil moisture patterns.

Project: Design and development of a robot for automated BLE-RSSI fingerprinting. • Designed and built an autonomous robot using Arduino, RaspberryPi, and BLE sensor to automate indoor signal mapping, to reduce human intervention in scene analysis.. • Developed a Python-based system for RSSI data collection, processing, and storage using Pandas and BluePy. • Implemented KNN-based fingerprinting for indoor localization, achieving 63% accuracy in predicting the position of the robot in the scene.

Project: Design and implementation of a modular solution for the automation of industrial radio-frequency tests in a Python environment with a graphical interface and with Raspberry Pi • Reduced manual RF testing time and improve operator safety. • Automated industrial RF tests using Raspberry Pi and Python. • Developed a multi-threaded GUI(Tkinter) with security automation for operators safety. • Traceability: Upload the necessary data of products to a server using SOAP requests.