portfolio

Automation & Robotics Engineering Portfolio | Low-level Embedded Systems, Edge AI, and Model Predictive Control (MPC).

Hi, I'm Szymon Markowski 👋

I am a 3rd-year Automation and Robotics engineering student at Poznan University of Technology, specializing in low-level embedded systems, non-linear optimization (MPC), and edge AI.

Note on my GitHub activity: The majority of my complex projects and daily coding are developed within the closed-source organization of my university's scientific club, PUT Powertrain. Below is a showcase of my key contributions, hardware designs, and algorithms.

---

🚀 Key Projects & Engineering Focus

1. Hardware Reverse Engineering & Custom Tooling (In Progress)

Integrating a closed-source motor into an autonomous boat without official technical documentation.

• What I'm building: A custom CAN bus protocol sniffer based on the RP2040 (Raspberry Pi Pico) and a dedicated hardware debugger using a Raspberry Pi 5 with a custom carrier board.
• The Goal: To capture raw data frames, manually decode the proprietary payload, map the state machine, and write a custom low-level C/C++ driver from scratch.

2. Autonomous Navigation & MPC Control

Developing the core movement logic for an autonomous boat operating in highly stochastic and noisy environments.

• What I did: Engineered a non-linear dynamic model and implemented a Model Predictive Control (MPC) algorithm to solve real-time constrained optimization problems for trajectory tracking.

• Description of my mathemacial work in polish is availible in the model_POWERBOATv2.pdf

3. Edge AI Vision Pipeline

• What I did: Applied transfer learning to fine-tune YOLO architectures for real-time object detection in marine environments. Deployed the deep neural networks on NVIDIA Jetson edge hardware to achieve high-throughput, low-latency inference.

4. PID temperature controller on STM

What I did:* Designed and programmed a custom PID control system for a ceramic resistor utilizing an STM32 microcontroller. Successfully integrated external peripherals, including a rotary encoder for dynamic target adjustments and a display module for real-time visualization.

• 🔗 Source Code: The complete codebase for this project is available in a dedicated public repository directly on my GitHub profile.

---

🛠️ Tech Stack Overview

• Languages: C++, C, Python, MATLAB
• Embedded & Automation: STM32, RP2040, TIA Portal, Simatic Manager, OPC UA, Raspberry PI 5
• Tools & Frameworks: ROS 2, PyTorch, TensorFlow, Git, Linux

📫 Reach me at: [Wklej swój link do LinkedIn] | markoszym2004@gmail.com