Dušan Grković

Technologies: LabVIEW, Python, OpenCV, MediaPipe, NI sbRIO, FPGA

• Real-time simulation and control of a serial two-joint manipulator featuring PID, Fuzzy-PID, and Sliding Mode Control; LabVIEW GUI with automated path tracking and Python/OpenCV/MediaPipe hand-tracking; validated on NI sbRIO/cRIO hardware. Repository
• Collaborated in a team to model the kinematics and dynamics of a two-joint robotic arm.
• Implemented and compared the performance of multiple control strategies, including PID, Fuzzy Logic-tuned PID, and Sliding Mode Control (SMC).
• Deployed and tested the control algorithms on NI sbRIO real-time hardware to validate performance.

Technologies: ESP32, STM32, ESP-NOW, BLDC Motors

• Designed a rover using ESP32 and STM32-based BLDC drivers with traction control, leveraging ESP-NOW for robust wireless communication.

Technologies: Siemens TIA Portal, Ladder Logic

• Completed practical projects applying Ladder Logic for industrial automation, demonstrating proficiency in PLC-based control.

Technologies: XMC4700, Arduino IDE, BLDC Motors, Fuzzy Logic, PID Control

• Developed a complete smart robotic grasping system from concept to execution in a fast-paced, 24-hour team challenge.
• Implemented a discrete PID controller to precisely manage the system's core control loop and motor actuation.
• Designed and integrated a fuzzy logic inference engine for real-time PID gain tuning, enabling the system to intelligently adapt its grip force.
• Engineered an algorithm to infer object hardness by processing real-time data from the hand's force sensors, providing the critical input for the adaptive grip.

Technologies: ESP32, Python, C++, HIL Simulation, Serial Communication

• Built a Hardware-in-the-Loop (HIL) testbench to develop and test an automated airlock control system for the European Rover Challenge 2025.
• Wrote C++ firmware for an ESP32 to operate a three-zone airlock, including its automated gates and safety sensors.
• Created a Python GUI to visually simulate the airlock in real-time for interactive testing.
• Set up serial communication for the simulator to exchange data with the ESP32 controller.

Technologies: Arduino, C, PWM

• Engineered a low-cost servo controller using a brushed motor and Arduino Uno, achieving precise motion control.

Technologies: ESP8266, C++

• Built a remote control system with integrated temperature and humidity sensors, enabling smart home automation.

Technologies: Intel 8051, C, Modbus RTU

• Implemented Modbus communication on an Intel 8051 microcontroller, ensuring reliable data exchange in resource-constrained environments.

Technologies: Python, OpenCV, Flask

• Created a web application using OpenCV to detect available parking spaces in real-time, optimizing urban mobility.

Technologies: Python, Scikit-learn, Pandas

• Developed a predictive model using machine learning to identify diabetes risk, achieving high accuracy on test datasets.

Technologies: Matlab

• This project provides a MATLAB script for calculating a range of financial technical indicators from stock market data. The script processes, normalizes, and structures this data into features and targets, designed to aid in market analysis or the development of predictive trading models.

Technologies: Android Studio, Java, Bluetooth

• Designed an Android app for sensor fusion, interfacing with a microcontroller via Bluetooth for real-time data processing.

Technologies: MATLAB, Signal Processing

• Compared FFT-based noise cancellation in MATLAB with analog filter designs, evaluating performance trade-offs.

Technologies: C, Image Processing

• Implemented an efficient image processing algorithm for edge detection, optimized for performance.

Technologies: Algorithmic Trading, WebSocket, JSON, JavaScript, Derivatives (Futures & Options)

• Collaborated in a team of four to design and build a high-frequency trading system from the ground up during the intensive 24-hour AlgoTrade 2025 hackathon.
• Engineered the system to execute strategies on futures and options by interfacing with a real-time, JSON-based WebSocket API.
• Achieved a critical low-latency advantage by deploying the trading logic on a collocated Raspberry Pi for near-instant order execution.
• Successfully managed a $1,000,000 virtual portfolio, applying robust risk management principles to maximize profit in a competitive simulated market.

Technologies: C, File Systems

• Wrote a C program to recover JPEG images from raw binary dumps, demonstrating low-level data manipulation.