Erik Mischiatti

Mechatronics engineer working across robotics, autonomous systems and system integration.

I build and test robot systems that connect embedded hardware, perception, control and autonomous decision logic for real-world prototypes.

Test EngineerMRS Group, FEE CTU in Prague

Hi! I'm Erik,

Mechatronics Engineer with a strong passion for robotics, automation, and intelligent systems. I hold a Master's degree in Mechatronics Engineering, majoring in robotics and electronics. My work mainly revolves around learning from demonstration, modular robot design, and autonomous task execution, especially in real-world, uncertain scenarios.

Alongside my academic path, I've gained experience in UWB-based drone communication, embedded systems, and CAD/PCB design. I love working at the intersection of hardware and software to build robust, adaptive technologies.

Beyond my technical interests, I enjoy sports like running and swimming, as well as analog photography. Originally from Italy, I'm currently living in Prague, where I work, and actively exploring new opportunities to develop both professionally and personally.

Featured projects

01

M.I.R.A.: Modular Interactive Robotic Agent

Personal, open sourceActive desktop prototype

A software-first embodied assistant built with Python and PySide6: an animated face driven by a renderer-independent embodiment model, modular cognition with safe local actions, and an ambient notch interface in development.

Built with
Python 3.12, PySide6 / Qt 6, pytest, GitHub Actions, optional Ollama integration
My contribution
Personal project, published on GitHub.
Evidence
Software implementation. Automated tests and CI cover cognition, actions, embodiment and UI behavior. Not a hardware robot yet.
Main window: idle micro-motions, then a typed exchange (replies are in Italian).

02

NEXUS: Modular Robotics Operations Platform

Personal, open sourceEarly implementation

A plugin-first operations platform for heterogeneous autonomous systems: a TypeScript core runtime and plugin platform, and a browser operator UI shell. Robot adapters are the next phase.

Built with
TypeScript, Node.js, React, Vite, Vitest
My contribution
Personal open-source project.
Evidence
Software implementation. Core runtime, plugin platform and UI shell are covered by automated tests. The UI runs on mock data; no robot is connected yet.
  1. Event bus
  2. Configuration, logging, services
  3. Runtime bootstrap
  4. Plugin platform
  5. Operator UI shell (mock data)
Implemented building blocks. Adapters, the robot model, capabilities and missions are the planned next layers.

03

Behavior Tree Mission Manager

EAGLE.ONE, FEE CTU in Prague, with Fly4FutureActive integration, since Oct 2025

ROS 1 mission-level supervisor for an interceptor UAV, built around modular BehaviorTree.CPP pipelines, mission-state reporting, service-gated runtime control, external target guidance, and ground-station/MQTT integration.

Built with
ROS 1 Noetic, C++, BehaviorTree.CPP, MRS UAV System, Gazebo simulation, GoogleTest
My contribution
Develop and test BehaviorTree.CPP logic for preflight, takeoff, pre-intercept, interceptor, post-intercept, landing, and finalization phases. (CV)
Evidence
Software implementation, simulation. Verification is layered. Simulation and software tests deliberately distinguish what is verified in software from what requires physical flight validation.
  1. Preflight
  2. Takeoff
  3. Pre-Intercept
  4. Interception
  5. Post-Intercept
  6. Landing
  7. Finalization
Mission phase sequence, as published on the project page. Each phase is an independent behavior tree.

04

UWB Communication System for Modular UAVs

EES Lab, University of TrentoCompleted, published 2024

Ultra-wideband links replace the wires between flight controller and ESCs, so rotor modules can attach to the payload itself.

Built with
Decawave DWM1001C, 6489.6 MHz channel, up to 6.8 Mbps; tri-, quad- and hexacopter layouts
My contribution
One of four student contributors. Third author of the IEEE CASE 2024 paper.
Evidence
Hardware prototype, field test, experimental validation, publication. Outdoor flight; hardware-in-the-loop tests show stable control around a 5 ms end-to-end delay.
3D model of a rotor module, with the UWB board in its printed case.
3D model of a rotor module, with the UWB board in its printed case.
Exploded CAD view of the three-part enclosure: antenna mount, battery section, electronics case.
Assembled base station with GNSS antenna dome, mounted on a tripod in a lab.
Exploded CAD view of the enclosure, and the assembled Version I on its tripod.

05

RTK Base Station

EAGLE.ONE, FEE CTU in PragueVersion I complete, Version II in progress

A portable GNSS reference station. It runs Survey-In at power-on and broadcasts RTCM corrections over LoRa, designed for centimetre-level RTK positioning of drones and rovers.

Built with
Here4 Base (u-blox F9P), LoRa E220 at 900 MHz, 4 × 18650 pack, three-part printed enclosure
My contribution
Hardware selection, enclosure design, field deployment. Enclosure adapted from an open design by Burks Builds.
Evidence
Hardware prototype. Field test, reported. Field use with drones is stated; no measurements are published.

Index

All projects and what each one actually produced, as documented on its project page.

Projects and their engineering evidence
Project Context Period Evidence
Hardware prototypeField testExperimental validationSimulationPublicationSoftware
RTK Base Station EAGLE.ONE, FEE CTU in Prague –Hardware prototypeField test, reportedNoNoNoNo
UWB modular UAV EES Lab, University of Trento 2024Hardware prototypeField testExperimental validationNoPublicationNo
BT Mission Manager EAGLE.ONE, FEE CTU in Prague 2025–NoNoNoSimulationNoSoftware
Cable manipulation, M.Sc. thesis TU Wien 2024NoNoExperimental validationNoNoSoftware
Green Vigilance – Baseline completeNoNoNoSimulationNoSoftware
EMG acquisition, B.Sc. thesis University of Padova 2021Hardware prototypeNoExperimental validationNoNoNo
VISTA, Behavior Tree layer Contribution to A. Ricciardi's ROS 2 framework In progressNoNoNoNoNoSoftware
M.I.R.A. Personal, desktop assistant prototype ActiveNoNoNoNoNoSoftware
Hachi, AR learning cards CLab Trento with HIT, team; DigiEduHack 2021 award 2021NoNoNoNoNoSoftware
NEXUS Personal, operations platform Early implementation, no robot adapters yetNoNoNoNoNoSoftware
3D prints Personal, published on Cults3D –NoNoNoNoNoNo

Documented in the project materialReported, no data published

Experience

  1. Oct 2025 – present

    Test Engineer

    MRS Group, FEE CTU in Prague

    EAGLE.ONE drone-hunting system, in collaboration with Fly4Future s.r.o.

Education

  1. 2024

    Master's thesis

    TU Wien

    Imitation learning and 6D pose estimation for task learning.

  2. 2022 – 2024

    M.Sc. Mechatronics Engineering

    University of Trento

    Focus on autonomous systems and robotics.

  3. 2023

    Erasmus exchange

    Linnaeus University

    Robotics, automation and mechanics.

  4. 2018 – 2021

    B.Sc. Mechatronics Engineering

    University of Padova

Skills and expertise

Robotics and autonomy

  • ROS
  • Behavior Trees (BehaviorTree.CPP)
  • Gazebo simulation
  • Control, perception and planning integration

Perception and learning

  • Learning from Demonstration
  • OpenCV
  • Object detection
  • Pose estimation

Embedded systems

  • STM32
  • Arduino
  • UART, SPI, I²C
  • UWB modules, sensors, motor drivers

CAD and PCB design

  • SolidWorks
  • Fusion 360
  • Onshape
  • EasyEDA

Software

  • Python
  • C++
  • MATLAB
  • Processing

Tools and deployment

  • Linux
  • Git (GitHub, GitLab)
  • Docker, Apptainer
  • ROS build systems

Publication and recognition

L. Santoro, A. Assirelli, E. Mischiatti, M. Perini, M. Nardello, D. Fontanelli, D. Brunelli. Deconstructing for Superiority: Aristotelian Modularization in UAV Systems. 2024 IEEE 20th International Conference on Automation Science and Engineering (CASE), pp. 219–224. doi:10.1109/CASE59546.2024.10711488

DigiEduHack 2021 Global Award, Hachi team. 228 submissions, 12 finalists. Announcement

Contact