Athanasios Martsopoulos

Robotics Engineer

Bristol, United Kingdom amartsop@gmail.com

I am a Robotics Engineer with six years of experience in both software and hardware development for aerospace and robotics applications. Skilled in C++, Python, and C. My interests and experience cover a wide range of tools and technologies, including mathematical modelling of dynamical systems, control systems design, state estimation and sensor fusion, system identification, trajectory optimization/path planning for robotic systems, high-performance computing, computer graphics and computer vision. I hold an MEng in Mechanical Engineering (First-Class Honours) and I am currently pursuing my PhD in Robotics at the University of Bristol.


Skills and Interests

Programming Languages & Tools
Interests

As a robotics engineer, I have worked extensively with robotic manipulators, ROS, perception, motion control and path planning algorithms. Examples of such activities include the development of motion tracking, 3D reconstruction and computer vision algorithms using OpenCV and depth cameras, development of SLAM, sensor fusion and state estimation algorithms as well as the and the design of control systems for linear and nonlinear applications (MPC, LQR, Sliding Mode, Feedback Linearization). Furthermore I have desinged and programmed fully autonomous robotics systems.

Apart from robotics software development, I have extensive experience with the mathematical modelling of physical systems. My interests mainly focus on computational mechanics including the dynamics of rigid, multibody and deformable systems. Over the years, I have used and developed software for solving, analyzing and visualizing such systems. Examples of these tools are numerical solvers for both ODEs and PDEs, efficient collision detection algorithms (e.g spatial hashing, KD-trees), rendering software (OpenGL) and constraint methods for contact handling.


Experience

Software Developer

SLiNK-TECH

  • Contributed to the development of a drone simulation environment in Unreal Engine 4.
  • Developed inter-process communication applications using UDP/TCP protocols and networking libraries (ZeroMQ).
  • Developed hardware-in-the-loop (HITL) and software-in-the-loop (SITL) simulations using embedded systems (ArduPilot/Pixhawk) and the MavLink communication protocol.
  • Used Docker to develop and deploy microservices architectures.

Oct 2022 - Present

PhD Candidate Robotics and Autonomous Systems

University of Bristol

My PhD focuses on creating high-fidelity mathematical models that aim to enhance our understanding of the physics of minimally invasive surgery and facilitate the development of robot-assisted surgery solutions. During my PhD, I:

  • Programmed multiple robotic manipulators (UR5, UR10, Franka Emika Panda, Schunk LWA4P) using ROS and MoveIt.
  • Implemented computer vision algorithms with OpenCV for real-time motion tracking and 3D reconstruction of deformable objects.
  • Developed an autonomous robotic system using ROS. The system integrated motion planning, com- puter vision, optimal control (MPC), state estimation and sensor fusion algorithms (Kalman filters).
  • Developed a real-time physics-based surgical simulation environment using C++ and OpenGL.
  • Developed efficient computational architectures and used parallel computing frameworks (CUDA, OpenMP).
  • Developed mathematical and computational tools, based on the theory of continuum mechanics and flexible multibody dynamics to allow the modelling of both soft human tissue and flexible surgical instruments.
  • Developed and implemented highly efficient numerical algorithms for the solution of PDEs and ODEs for real-time applications.

Sep 2019 - Present

Research Associate in Robotics

University of Bristol

  • Developed a HITL simulation using Arduino and OpenGL.
  • Developed motion tracking algorithms using machine learning and depth cameras.

Nov 2021 - Mar 2022

Research Associate in Flight Mechanics

CIRI - Aristotle University of Thessaloniki

  • Contributed to the development of a prototype VTOL UAV.
  • Developed mathematical models of UAVs.
  • Designed implemented and tested navigation and motion control algorithms.
  • Worked with embedded systems programming and hardware-in-the-loop simulations.
  • Conducted extensive testing and software development in C/C++, Python and MATLAB.

Dec 2018 - Sep 2019

Flight Dynamics and Control Engineer

Aristotle Space & Aeronautics Team

  • Contributed to the design and sizing of UAVs for Air Cargo Challenge (Aeronautical Engineering Competition).
  • Evaluated the static and dynamic stability UAV characteristics.
  • Implemented flight simulations using FlightGear and MATLAB.
  • Designed, constructed and programmed measuring devices for the experimental evaluation of mathematical models.
  • Trained new team members.
  • Oct 2016 - Dec 2017

    Education

    University of Bristol

    PhD Robotics and Autonomous Systems

    Development of mathematical and computational tools for the real-time simulation of prostate biopsy/brachytherapy surgery.

    Sep 2019 - Present

    Aristotle University of Thessaloniki

    Diploma in Mechanical Engineering (MEng)

    First-Class Honours, Valedictorian : 8.97/10

    • Specialization Field : Design and Structures
    • Level of Qualification : Integrated 1st and 2nd Cycle
    • Official Length of Programme : 10 Semesters, 300 ECTS
    Sep 2013 - Sep 2018

    Publications

    ...
    Master's Thesis

    Nonlinear control of an autonomous flying wing aerial vehicle

    ...
    IROS 2020, Conference Paper

    Spatial Rigid/Flexible Dynamic Model of Biopsy and Brachytherapy Needles Under a General Force Field.

    ...
    MCMDS, Journal Paper

    Modelling and real-time dynamic simulation of flexible needles for prostate biopsy and brachytherapy

    ...
    ICRA, 2023

    Development and Experimental Verification of a 3D Dynamic Absolute Nodal Coordinate Formulation Model of Flexible Prostate Biopsy/Brachytherapy Needles

    Personal Projects

    Robotic arm simulation and control

    As part of my PhD I designed and developed an experimental setup for controlling Schunk Lwa4p robotic manipulator and tracking/controlling flexible medical needles, inserted into soft tissue. The illustrated ImGUI dashboard allowed the user to control the robotic manipulator, visualize its forward kinematics(ROS/MoveIt) in 3D using OpenGL and track in real-time the flexible needle using 3D reconstruction from two high frame-rate cameras.

    Real-time simulation of deformable human tissue

    As part of my PhD, I have developed multiple algorithms for the real-time/interactive simulation of deformable bodies, focusing on human tissue. For this I have used a wide range of techniques for continuum mechanics and graphics theory, including Finite Element Methods (FEM), Corrotational FEM, Meshless approaches (Radial Point Interpolation, Galerkin), Mass-Spring Systems (MSM) and Position Based Dynamics (PBD and XPBD). The simulation environment was implemented from scratch with the help of C++ and OpenGL.

    Real-time simulation of flexible medical needles

    This application illustrates the implementation of a real-time flexible needle models. These models, based on the theory of multibody dynamics (etc. RFEM, ANCF), allow the incorporation of geometric nonlinearities, large deflections and arbitrarily large rigid body motions. Furthermore, they are characterized by high computational efficiency and accuracy.

    Collision detection, handling and interaction mechanisms for deformable bodies

    This application illustrates the collision detection and handling algorithms I developed as part of my medical simulation environment.

    Hand exoskeleton

    This hardware-in-the-loop simulation allows real-time rendering of the three-finger exoskeleton with the help of a photorealistic hand model. The proposed model is based on the principles of skeletal animation.

    Inverted double cart-pendulum simulation

    This project illustrates the simulation of a double pendulum on a cart, using the SFML graphics libary and Armadillo (linear algebra).

    Inverted cart-pendulum and nonlinear model predictive control (MPC)

    A nonlinear implementation of model predictive control for swing-up and stabilization of the cart-pole in the upper equilibrium point.