About Me Education Work Experience Selected Papers Selected Projects

About Me

Hey, I'm Marcos Moura

I'm Marcos, a Brazilian, passionate about samba šŸŖ˜ robotics šŸ¤– programming šŸ‘Øā€šŸ’» AI šŸ§  and more šŸš€
I currently work as a Principal Engineer at the University of South-Eastern Norway in the fields of Robotics, Computer Vision, Software Development and Electronics.
To find out more about my projects, have a look at the other sections, or at my full Resume.

View my Resume

Education

Master in Sensor Systems and Innovation

University of South-Eastern Norway

Horten - Norway

Aug. 2025 - Present

In this master course I'm diving into Sensor Technologies and how to integrate them in real-world challenges.

Lato Sensu Postgraduate Degree - Specialization in Full Stack Development

Puc - Minas

Remote

Aug. 2024 - Aug. 2025

In this specialization course I learned several technologies for Full-Stack development for web and mobile, AGILE project management, DevOps practices.

BSc in Electronic Engineering

CEFET/RJ

Rio de Janeiro - Brazil

Feb. 2016 - Mar. 2022

My studies here marked the beginning of my professional and scientifical path. In parallel with my studies, I acted in a project team that develops model aircrafts for an annual national competition, more information about it can be found in the Venturi Aerodesign' LinkedIn. I also acted as a student researcher in the fields of Photonics and Robotics in the Laboratory of Photonics of CEFET/RJ and in the Magnetic Resonance Laboratory of the Brazilian Centre of Physical Research (CBPF). My bachelor thesis can be found here.

Technician in Industrial Automation

SENAI

Duque de Caxias - Brazil

Feb. 2014 - Dec. 2015

My first steps in the world of robotics was in this vocational school. Here I learned the basics of electronics, automation and industrial instrumentation, which gave me valuable insights for the next steps.

Work Experience

Principal Engineer

University of South-Eastern Norway

Full-Time | Presential

Jul. 2025 - Present

I work for the Department of Microsystems at USN, and between my roles I have to: Maintain and fix electronic equipment at the robotics laboratory; write and maintain code to implement algorithms, perform simulations, and interface with robotic hardware platforms; contribute to EU Horizon and NFR projects in which the Department of Microsystems is involved; conduct literature reviews and document research methodologies, experimental procedures, and results; present findings and results through presentations, reports, and possibly publications, both internally and at conferences

Scientific Assistant in Robotics and Machine Learning

University of South-Eastern Norway

Full-time | Presential

Jul. 2023 - Jul. 2025

In my first position at USN, I assisted researchers in designing, conducting, and documenting experiments related to robotics and machine learning projects. I also wrote and maintained code to implement algorithms, perform simulations, and interface with robotic hardware platforms, conducted literature reviews and documented research methodologies, experimental procedures, and results and presented findings and results through presentations, reports, and possibly publications, both internally and at conferences

AD&ADAS System Design Engineer, Exterior Vision

Volvo Cars

Full-time (Temporary) | Presential

Jan. 2023 - Jul. 2023

Working as a System Design Engineer, I verified the exterior cameras algorithms delivered by suppliers and reviewed camera system requirements according to ISO26262.

Embedded Systems Developer

TracSense AS

Full-Time | Presential

Nov. 2021 - Dec. 2022

In my first job in Europe, I was responsible for designing the embedded systems with microcontrollers of the STM family. I also integrated sensors in ROS and developed camera drivers for Linux and HMI softwares.

Junior Web & Sensor Developer

TracSense AS

Full-Time | Remote

Sep. 2020 - Nov. 2021

I was responsible for programming microcontrollers of the STM and ATmega family. I also assisted with the development of RESTfull APIs and frontend applications for the AutonoWeather Project

Full-Stack Developer

Danke Delivery

Part-time | Remote

Jun. 2021 - Dec. 2021

I was responsible for developing two web-apps in React, hosted on Firebase, and for the backend integration. This was a start-up that I developed with some friends. From this experience, I gained knowledge of software development, web frameworks, and AGILE management.

System Development Intern

ECOA - Puc/Rio

Internship | Remote

Feb. 2020 - Oct. 2020

My first opportunity as a programmer was in the Ignition Program, a cooperation between Puc-Rio and Petrobras. In this program, we were presented with some of the issues in the Brazilian oil industry and had to organize ourselves to develop MVPs and present solutions for the given problem. It was my first contact with web development and AGILE management.

Selected Papers

Optimizing Object Detection for Maritime Search and Rescue: Progressive Fine-Tuning of YOLOv9 with Real and Synthetic Data

In this work, in which I was a co-author, we keep exploring how mixed datasets can be used for transfer learning. Here, the use of real and synthetic data for training the model YOLOv9 in maritime search and rescue operations is explored. The technique used for transfer learning was fine-tuning. Different proportions of real data were used, to evaluate the performance aiming to reduce the reliance on real-world datasets.

Full Paper →

Road Weather Forecasts in Norway with the METRo Model

In this paper, in which I worked in the API development and data curation, we evaluate road weather forecasts in Norway using the METRo model in a quasi-operational setup. Initialized with road weather station data and mesoscale forecasts, the model focuses on predicting freezing road-surface temperaturesā€”a key hazard in Norway. By comparing hit and false-alarm rates to climatology, we show that METRo adds consistent skill, especially in winter, spring, and fall. This study demonstrates a feasible, open-source approach to operational road weather forecasting using accessible data.

Full Paper →

Development of a Solar Panel Control Strategy for Tracking Maximum Power Generation

In this work, in which I was a co-author, we describe the prototype of an electronic supervision and control system for the orientation of a bench solar panel. The developed tracker prototype has as its core an electronic circuit based on a commercial microcontroller model Tennsy 3.0, within which the control algorithm is embedded. In addition to the controller, a supervisory software was developed to monitor solar cells' status in real-time. The supervisory showed the angle of the solar plate and values of luminosity and acquired power. Simulations results were presented to show that the amount of energy generated can reach 37 %.

Full Paper →

Empowering Search and Rescue Operations with Big Data Technology: A Comprehensive Study of YOLOv8 Transfer Learning for Transportation Safety

In this project, in which I was co-author, we explores how object detection, optimized through transfer learning and synthetic data, can improve maritime search and rescue missions using unmanned aerial vehicles. Due to limited real-world sea operation data and onboard processing constraints, the use of synthetic datasets and transfer learning offers a flexible solution. YOLOv8 models were trained with various configurationsā€”using real, synthetic, and mixed datasetsā€”under both regular and transfer learning setups. Performance was evaluated using mAP50-95, highlighting key differences between synthetic and real data in training outcomes.

Full Paper →

Reflectance and transmittance spectra of a photonic crystal slab with doped graphene disks for operation at the THz band

This conference paper, in which I was main author, we present seminal results to characterize the performance of a two-dimensional photonic crystal slab waveguide with doped graphene disks for operation at the Tera-hertz spectrum. More specifically, we analyze the reflectance and transmittance spectra for frequencies ranging from 10 THz up to 20 THz. It is shown that a broader transmission spectrum can be achieved when compared to the analogous all-dielectric structure, suggesting promising applications for filtering performance at the THz domain.

Full Paper →

Design of a surface plasmon resonance flip flop S-R circuit

In my bachelor thesis, me and my supervisor explores the propagation characteristics of plasmonic modes to design an optical version of an S-R flip flop, a sequential digital circuit that can be used as a one-bit memory device. The performance of the structure and its operation as an SR flip flop is analyzed in terms of simulation results for the transmission spectrum, considered as the square magnitude of the S parameters (|Sij|2) between a pair of specified input-output.

Full Paper →

View more papers

Selected Projects

Firebase Functions Template

A Typescript template for Firebase backend functions to make a CRUD in database. Since I had to set-up a backend in Firebase several times, I ended up coding this template to make it faster in the next times.

Source code →

Direct Step Edge Follower

I'm part of the team who developed the DSEF Python library.The Direct Step Edge Follower (DSEF) is a edge-following algorithm designed for high-precision edge detection with low computational cost. It employs stepwise directional refinement and kernel-based statistical testing to enhance accuracy, particularly in challenging lighting conditions.

Source code →

SLAM in ROS1

This project simulates a robot that uses opencv to find the exit from a room. The robot used, the P3DX, has a camera and a laser scanner. In this project, the robot rotates around the Z axis until it finds the door, then follows the path there. It was one of my first contacts with robotics, in 2020, during the pandemics

Source code →

Flask-ROS2 Video Stream

In this project, I unite two of my passions: web development and robotics. This is a ROS2 package for a Flask based web-app for video stream.

Source code →

Optical Pyber

A Package with tools to analize waveguides and optical fibers. This package uses numerical methods to find the eigenvalues related to the propagation modes and other characteristics of the waveguide. Currently, this package supports three types of waveguide: Planar, Retangular and Optical Fibers. This was developed during my studies in applied electromagnetism in my bachelor.

Source code →

PID Maximum Power Tracking

Implementation of a PID control for tracking the maximum power generation of a solar panel. The system is consisted of 2 LDRs, a Teensy 2.0, with a 32-bit ARM Cortex-M4, SG90 servo and the HMI was made using Elipse Scada.

Source code →

For more projects, click here