Amine Halfa

I am an embedded systems engineer with over 7 years of experience in low-level C/C++ development, specializing in STM32, ESP32, and nRF52 microcontrollers, in real-time environments (bare metal, FreeRTOS, Zephyr).

My expertise covers the entire embedded chain:

Firmware design and development for critical and low-power systems

Sensor, ADC, I2C, SPI, UART, PWM control

Wireless communication: Bluetooth Low Energy (BLE), Wi-Fi, LoRa

Embedded Linux integration, buildroot / Yocto / Device Tree

I have worked on projects:

🏠 Home Automation & IoT:

Development of firmware for connected objects for smart homes and elderly assistance. Multi-sensor reading (activity, falls, heart rate), BLE/Wi-Fi communication, and remote monitoring. Automation, mobile interfacing, and integrated data security.

🩺 Medical & Connected Health:

Design of embedded medical devices for monitoring medication intake and analyzing physiological data. Integration of biomedical sensors, real-time processing, secure transmission, user ergonomics, and compliance with health constraints.

🤖 Mobile Robotics:

Development of real-time firmware for autonomous mobile robots. Motor control, 360° perception (ToF sensors), inertial data fusion, audio processing (BF, EC, SL), 3D SLAM with ROS, Jetson Nano, and Linux integration. Integrated smart charging stations.

🧠 Embedded AI & Intelligent Assistants:

Prototyping voice interfaces on ESP32 and STM32 with embedded AI models (Edge Impulse, TensorFlow Lite). Event detection, voice activation, LED/audio feedback, OpenAI API, real-time local processing on memory card, Bluetooth, and I2S.

I am capable of delivering robust, documented, and optimized firmware, participating in electronic integration, and communicating with hardware, QA, and mobile teams.

BLE firmware development on nRF52 MCUs (832, 840)

Firmware development for a connected medical device for medication intake monitoring for the elderly. Technologies used: BLE, RGB LED, Visual Studio Code, Zephyr RTOS, nRF52, embedded C/C++.

Development of a 360° audio conferencing system integrated into mobile robots. Implementation of 3D SLAM (localization and mapping) algorithms under ROS. Technologies: Jetson Nano, D435i camera, Nvidia GPU, CUDA, embedded Linux, Python, C++, Respeaker V2.

Porting a road traffic data collection application (C++/OpenCV) initially developed for rugged PCs to a low-cost embedded Linux platform Apalis TK1.

Using Yocto to generate an embedded Linux image including necessary libraries (Qt, OpenCV). Development and integration of a graphical user interface (GUI) under Qt for real-time visualization of the IDS camera stream and traffic statistics (detection, counting, vehicle classification, speed estimation). Adaptation and optimization of existing code to ensure smooth operation on embedded hardware. Utilization of pre-trained AI models for vehicle detection and classification in the adapted application. Multithread management with QThread to ensure fluidity between camera acquisition, OpenCV AI processing, and Qt/QML display. Technologies: C++, Qt, OpenCV, Yocto, Embedded Linux, Ixora board, TK1 COM, IDS Camera, OpenMP, AI, Deep Learning, Websocket, Python, HTML/CSS/JS