Miroljub Mihailovic

Vicenza - Italy - miroljubmihailovic98@gmail.com

I am a dedicated individual with a strong background in Artificial Intelligence and Robotics. I hold a bachelor's degree in Computer Science and a master's degree in Artificial Intelligence and Robotics from the University of Padua. Currently, I am a PhD Student in Robotics and AI at UniPd, working at the DEI - IAS LAB. My research focuses on developing a new control strategy for lower limb exoskeletons, emphasizing shared autonomy between human and robot intelligence. Proficient in Python, Java, and C++, I have experience in various AI and robotics projects. As a team player with strong problem-solving skills, I am committed to staying current with the latest technologies and always seeking opportunities to expand my knowledge.


Experience

ML Researcher

Nordest Technology

Machine learning researcher providing time-limited support for the development of algorithms and models for detecting anti-money laundering (AML) activities.

11/2024 - Present

Lead Software Engineer in Machine Learning

Nordest Technology

As a Lead Software Engineer in Machine Learning, my core responsibility revolves around leading the development efforts of AI and ML-based products specifically designed for the detection of Anti-Money Laundering (AML) activities.

01/2023 - 11/2024

Software Engineer

DEI - IAS LAB (University of Padua)

Software engineer working on refactoring code for ROS-Neuro system.

02/2024 - 06/2024

Mentee at LeadTheFuture

Lead The Future Mentorship

Among the few Italian students selected to be mentees for LeadTheFuture, a leading mentorship non-profit organization for students in STEM, with acceptance rate below 20%. LeadTheFuture empowers top-performing students to achieve their goals and contribute to their communities by giving them one-on-one guidance from high-impact mentors coming from the world's leading STEM innovation hubs such as Silicon Valley (Google, Microsoft, Facebook..) and CERN.

09/2022 - Present

Software Engineer in Machine Learning

Nordest Technology

In this role, I've been instrumental in developing software to optimize the identification of suspicious financial transactions. Utilizing machine learning, particularly supervised learning, our software ranks and prioritizes Anti-Money Laundering (AML) alerts, reducing false positives and enabling efficient alert handling for enhanced financial security.

04/2022 - 01/2023

Software Engineer

ESA (European Space Agency), DII (Department of Industrial Engineering University of Padua).

This project is concerned with the design and development of a test for an autonomous docking manoeuvre between CubeSats mock-ups to be performed on a parabolic flight in order to take advantage of a reduced-gravity environment. Co: Bortotto A., Degli Agli G., Favotto F., Mattiazzi F., Pozzato N.

12/2021 - 11/2022

Research Engineer

DEI - IAS LAB (University of Padua)

I investigated how to integrate reinforcement learning in a shared intelligence system where the user’s commands are equally fused with the robot’s perception during the teleoperation of mobile robots. Specifically, I developed a new policy-based implementation suitable for navigating in unknown indoor environment with the ability to avoid collisions with dynamic and static obstacles. The aim consists of extending the current system of shared intelligence based on numerous pre-defined policies with a new one based on Reinforcement Learning (RL).

10/2021 - 07/2022

Education

PhD Student in Robotics and AI at UniPd

DEI - IAS LAB (University of Padua)

The research project aims to develop a new control strategy for lower limb exoskeletons, introducing a shared autonomy paradigm between human intelligence and the robot's artificial intelligence. Key objectives include modeling human movement using EMG and kinematic data, adaptively generating steps in complex environments, and creating a semi-autonomous control system that integrates user and exoskeleton capabilities.The research project aims to develop a new control strategy for lower limb exoskeletons, introducing a shared autonomy paradigm between human intelligence and the robot's artificial intelligence. Key objectives include modeling human movement using EMG and kinematic data, adaptively generating steps in complex environments, and creating a semi-autonomous control system that integrates user and exoskeleton capabilities.

11/2024 - 11/2027

University of Padua

Master of Science
COMPUTER ENGINEERING - ARTIFICIAL INTELLIGENCE AND ROBOTICS

GPA: 28.621/30

10/2020 - 07/2022

University of Padua

Bachelor of Science
COMPUTER ENGINEERING

GPA: 26/30

10/2017 - 09/2020

Skills

Programming Languages
Main Tools
  • cplex

Technologies

Main Projects

Code Refactoring for the ROS-Neuro. ROS-Neuro Code

My focus lies in two critical areas: executing unit tests and performing code refactoring for ROS-Neuro. These concerted efforts are aimed at bolstering reliability, validating functionality, and optimizing the codebase for enhanced maintenance and scalability.

Anti-Money Laundering Anomaly Detection. Kassandra - Prioritas

I spearheaded the development of Prioritas and played a pivotal role in overseeing and contributing to the creation of Kassandra. Prioritas is a cutting-edge system that speeds up the generation of Suspicious Activity Reports (SOS) by ranking AML alerts based on their potential significance. This innovation reduces SOS processing time by 60% to 80% and minimizes errors in predicting low-risk alerts. Kassandra, a powerful AML module fully based on artificial intelligence, uses pattern recognition to uncover hidden anomalies in transaction data. It boasts high precision, with 90% of subjects flagged for further investigation and 52% resulting in SOS reports. Together, these solutions have greatly enhanced our AML and transaction monitoring capabilities, ensuring efficient and accurate operations in the evolving landscape of financial security and compliance.

Experimental Rendezvous in Microgravity Environment Study (ERMES)

  • 4thSymposium on Space Educational Activities (Barcelona) - 2022. Abstract
  • 72nd International Astronautical Congress (Dubai, UAE) - 2021. Paper

  • Within the project, I played a crucial role in designing and implementing the Onboard Computer System (OBCS) of the CubeSat. The primary Raspberry Pi processes sensor data (tri-axial accelerometer, gyroscope, tri-axial magnetometer, proximity sensor) and computes trajectories for maneuvers, sending commands to the actuators. The secondary unit, based on Arduino, provides real-time control of the actuators. We utilized the Robot Framework and the Robot Operating System (ROS) to manage data and maneuvers. My contribution ensured reliable control of the onboard sensors and actuators, contributing to the success of the satellite missions.

    Reinforcement learning in shared intelligence systems for mobile robots. Thesis Code

    This thesis focuses on integrating reinforcement learning into a shared intelligence system for teleoperating mobile robots. The aim is to enhance the existing rule-based system with a reinforcement learning policy that enables the robot to navigate unknown indoor environments while avoiding obstacles. An agent learns through trial-and-error interactions, guided by a reward function inspired by the Attractive Potential Field. The robot's state is determined using a pre-processing module that clusters nearby obstacles. Various clustering algorithms are evaluated for real-time suitability. Different model configurations are tested in simulations, assessing the agent's reactive navigation with static and dynamic obstacles. The shared system, with the reinforcement learning policy, is compared with the current version in a teleoperated experiment, where high-level commands are issued to the robot for evaluation.

    Machine Learning project for boat detection. Code Report

    This university project outlines the implementation of a boat recognition system using machine learning and OpenCV techniques. The key components include a neural network for boat detection, a dataset generated from labelled images, and selective search algorithms to identify potential boat regions. A preprocessing phase enhances image quality and reduces the number of proposed regions. The detection phase involves analysing each region using the neural network and applying non-maximum suppression for efficient grouping. The project's results demonstrate the successful identification and bounding of boats in various images, showcasing the system's potential for boat recognition.

    Small thesis about different algorithms for solving the Traveling Salesman Problem (TSP). Thesis

    The work focuses on solving the Travelling Salesman Problem (TSP) using various techniques implemented in the C programming language. The approaches span from optimal solutions based on compact models, such as the Miller, Tucker, and Zemlin models, to methods inspired by Dantzig, Wilkerson, and Johnson, like the Benders method and CPLEX Callback. Additionally, a range of heuristics have been examined and compared, including Refinement Heuristics (Matheuristics), Constructive Heuristics, and metaheuristics like Variable Neighbourhood Search (VNS), Tabu Search, and Genetic Algorithm. This diversity of approaches has been employed to tackle the TSP in different ways.

    Awards

    • 2 rd Place - Vicenza - NASA Hackathon Space Apps challenge 2018 (Prize €1000)
    • 2 rd Place - Vicenza - NASA Hackathon Space Apps challenge 2017

    Interests

    My interests include reading, playing chess, sports, bodybuilding, running, and reading scientific articles on innovative topics. I enjoy playing basketball, football, and other games. Bodybuilding is a passion of mine, and I spend a lot of time working out. Running is also essential as it helps me stay in shape and clear my mind. Lastly, I am always interested in staying up-to-date on the latest scientific advancements and enjoy reading articles on innovative topics such as AI and Robotics.