Nabil Miri Portfolio

About Me

I'm Nabil Miri, a motivated Robotics Engineer with a passion for developing innovative robotic systems through research and creative problem-solving. My expertise lies in Model Predictive Control (MPC), mobile robotics, manipulators, and deep learning, enabling me to design and optimize robotic solutions for real-world challenges.

During my time as a Robotics Engineer Intern at BMW Group in Munich, I had the opportunity to work on cutting-edge projects that automated complex tasks in robotics. I developed connectivity features using Python, ROS2, and Docker, designed CI/CD pipelines, and integrated Azure services for seamless deployment across robot instances.

Recently, I worked on my Master's thesis at Fraunhofer IPA, where I developed a traversability-aware local MPC path planner for outdoor agricultural mobile robots. This work focuses on optimizing robot movement in challenging environments, leveraging advanced control techniques to improve performance and efficiency.

I have had the privilege of publishing two papers in IEEE conferences during my bachelor's degree. These publications, along with my role as a student researcher in the Robotics Systems and Technologies (RST) Department at TU Dortmund University, have provided me with a solid foundation in research, academic writing, and collaborative problem-solving.

I am committed to continuous learning and improvement, always striving to enhance my knowledge and skills. I thrive in collaborative environments where teamwork, communication, and dedication are essential to solving complex problems and advancing robotics technologies.

I'm passionate about leveraging my skills to develop innovative solutions that push the boundaries of robotics and intelligent systems.

Technical Skills

Programming Languages

Python
C++
MATLAB

Robotics & Control

ROS2
MPC
Navigation
Localization
Path Planning

Machine Learning & Vision

Deep Learning
3D Computer Vision
TensorFlow
PyTorch
Segmentation

Cloud & DevOps

Azure (Key Vault, Blob Storage, Table Storage).
Docker
CI/CD
Jira
Streamlit

Fraunhofer IPA - Master Thesis

Nov 2024 – Apr 2025

Stuttgart, Germany

Developed a Nonlinear Model Predictive Path-Following Controller (MPFC) for traversability-aware local planning in outdoor mobile robots.
Integrated continuous traversability costmaps into the MPC framework for optimal, context-aware navigation.
Validated on CURTtrac agricultural robot in both simulation and real-world field experiments.

Enable efficient local path following for outdoor agricultural robots in environments with irregular terrain and dynamic obstacles.
Balance path adherence, risk avoidance, and real-time performance on embedded hardware.

Designed a Model Predictive Path-Following Controller (MPFC) as a local planner within the Nonlinear Model Predictive Control (NMPC) framework.
Integrated continuous, differentiable traversability costmaps (from elevation data) and obstacle avoidance directly into the OCP.
Explored multiple OCP formulations: Traversability-Cost, Rate-Input, and Traversability-Constrained OCPs.
Solved using ACADOS framework (C-based, real-time optimal control) with SQP and RTI schemes.

Developed a ROS 2 / Nav2 local planner node for seamless integration with existing navigation stacks.
Tested in Python-based simulation (rapid prototyping), Gazebo (high-fidelity), and real robot experiments on CURTtrac.

Generated smooth, stable velocity commands and consistently selected low-cost traversable routes.
Operated at real-time control rates (20–33 ms on hardware).
Effectively avoided untraversable terrain and balanced traversability with goal achievement.
Demonstrated robust, context-aware navigation in complex field scenarios.

CURTtrac outdoor tracked robot (Fraunhofer IPA)

Experiment: Traversability-aware MPC demonstration

Thesis Presentation ARA Robot Project Website

BMW AG - Robotics Intern

Apr 2024 - Sep 2024

Munich, Germany

Developed communication protocols using Kafka and REST APIs for Smart Robotics Platform
Built CI/CD pipelines and cloud-based configurations for robotic deployments
Integrated ROS2 with cloud services using Docker containers

📍 Location: Munich, Germany | 🗓️ Duration: April 2024 – Sep 2024

During my internship at BMW Group, I had the privilege of working on cutting-edge robotics projects while collaborating with an exceptional, forward-thinking team. This journey allowed me to immerse myself in complex real-world challenges, contribute to innovative solutions, and grow both technically and personally. Being part of BMW’s robotics innovation ecosystem has been an inspiring and enriching experience. Worked in TV-204 under the amazing supervision of Marina Ezcurra Lopez-Roldan, with valuable support from Wafic El-Ariss, Fabio Falezza, and Daniel Jess.

Leveraged modern communication protocols to enable multi-robot configurations.
Hands-on impact in developing modular, scalable solutions for robotics use cases.
Learned the value of effective teamwork, iterative development, and clear documentation to deliver high-quality solutions.

🤖 Worked with a UR10e robot, integrating Kafka and RPCJson for communication protocols.
🔄 Refactored the Connectivity Package repository to improve modularity and enable the addition of new protocols like REST APIs and RabbitMQ.
☁️ Designed cloud solutions using Azure (Key Vault, Blob Storage, Table Storage).
🐳 Utilized Docker to create containers and automate launch/debugging processes.
🚀 Built and tested CI/CD pipelines for automated formatting, linting, and deployment.
🔧 Troubleshot and debugged ROS2 and Python errors for robot functionalities.

🗂️ Documented processes using Confluence and Markdown.
🤝 Worked in a Scrum-based Agile team, tracking progress via Jira.
🌱 Provided onboarding support and knowledge sharing for new team members.

TU Dortmund - Research Assistant

Apr 2023 – Mar 2024

Dortmund, Germany

Conducted research in Human-Robot Interaction (HRI), focusing on advanced motion planning techniques.
Collected RGBD data with Azure Kinect and trained NERF models using PyTorch.
Programmed robot manipulators and migrated ROS packages from Melodic to Noetic.

Contributed to cutting-edge research in Human-Robot Interaction (HRI), focusing on advanced motion planning, data collection, and system programming. Collaborated with the RST department to explore innovative robotic solutions, combining theory and application.

🧪 Assisted in testing in-house state-of-the-art Moving Horizon Planning for HRI, MHP4HRI, and replicated related research for comparative analysis.
📊 Collected RGBD data of robot environments using Azure Kinect and trained Neural Distance Fields (NDF) models with PyTorch using LocNDF library.
🔧 Programmed robot manipulators using ROS and C++.
🔄 Converted ROS packages from Melodic to Noetic to ensure compatibility and enhance system functionality.
📚 Maintained comprehensive documentation and user guides, and managed version control using GitHub.
👨‍🏫 Supervised by Heiko Renz.

🤝 Human-Robot Interaction (HRI)
📊 Azure Kinect
💻 PyTorch
🦾 ROS (Melodic/Noetic)
🔧 C++

Research Internship

May 2022 – Oct 2022

Troyes, France

🔬 Researched Bluetooth Low Energy (BLE) Protocol communication
💡 Programmed RN4020 BLE Module to create private services, implement MLDP mode, and manage data transmission using PIC microcontroller
🛠️ Focused on Design Optimization based on component calculations and achieving maximum possible outcomes
👨‍🏫 Supervised by: Dr. Aly Chkeir

Research Internship [Remote]

Jun 2022 – Sep 2022

Remote

🔬 Designed and Simulated an Affordable Potentiostat for electrochemical measurements
💻 Utilized Simulation Software to model circuits and design software like Microcap
👨‍🏫 Supervised by: Dr. Jana Abou-Ziki & Dr. Rolf Wuthrich

Education

Oct 2022 - Apr 2024

Dortmund, Germany

M.Sc. Automation and Robotics

🎓 Key Highlights

🧪 Research Assistant at RST (Lehrstuhl für Regelungssystemtechnik)
🏅 3rd Place Winner in TU Dortmund Startup Weekend
🚗 Group Project: Optimal Control of Chronos Car

Research Assistant at RST (Lehrstuhl für Regelungssystemtechnik)
Contributed to cutting-edge robotics research
TUD Startup Weekend 2023
3rd Place Winner - Developed an IoT-enabled smart cat collar concept
Group Project: Optimal Control of Chronos Car
Applied Reinforcement Learning (DQN and DDPG) for autonomous navigation
NMPC Course Project
Implemented Non-Linear Model Predictive Control for 2-DoF robot

German CGPA: 1.28

🤖 Robot Manipulators
🚗 Mobile Robotics
📸 3D Computer Vision
🧠 Machine Learning in Robots
📊 Nonlinear Model Predictive Control (NMPC)
🎛️ Optimal Control

Model Predictive Control
Deep Learning
Navigation
Robotics
Control Systems
Optimal Control
Mobile Robotics
Automation
ROS
Computer Vision

Aug. 2022 - Nov. 2022

Remote

Professional Certificate - Machine Learning

16-week intensive online training program in fundamentals of ML & DL designed to equip participants with market-ready skills

Final Project: Developed an Instance Segmentation Model (Mask-RCNN) for litter detection in the wild using Detectron2 lib and deployed it on the cloud using Streamlit.

View Litter Detection Project

2018 – 2022

Beirut, Lebanon

B.Sc. Mechatronics, Robotics, and Automation Engineering

GPA: 92.65%

Final Year Project:

Slug Picking Robot

🎓 Key Highlights

🚗 Final Year Project: Autonomous Mobile Robot for Outdoor Navigation and Slug Detection
1st Place: Beirut AI & Microsoft AI Bootcamp
2nd Place: Beirut AI University Hackathon
📜 RHU President's Honor List (All Semesters)

🎓 Grade: 92.65%

🏛️ Institution: Rafik Hariri University (RHU)

🚗 Developed an industrial-sponsored autonomous mobile robot using ROS. Implemented autonomous navigation algorithms and integrated YOLOv5 for slug detection and collection. Research published in IJMERR.
📸 Integrated YOLOv5 for detecting and collecting slugs objects. Manual data collection and labelling using labelimg.
🤖 Programmed and implemented navigation algorithm using MarvelMindGPS in Robot Operating System (ROS).

1st Place in Beirut AI & Microsoft AI Bootcamp (2020):
- Project: Esmaani – Software for aiding people with hearing impairments. Deployed on mobile using Tensorflow Lite.
- Issued by: Beirut AI & Microsoft.
2nd Place in Beirut AI University Hackathon (2020):
- Project: Developed an adaptive traffic system using YOLOv3 to detect and count cars waiting at traffic and prioritize roads.
- Theme: AI for a self-sustained Lebanon.
- Issued by: Beirut AI.
RHU President's Honor List in all semesters.

ASME Student Regional Vice Chair (Middle East & Africa):
- Connected and developed ASME Student Sections in the MENA region.
- Organized multiple seminars and linked members to professionals.

Published 2 research papers in the field of electronics and robotics.

Despite being a long and challenging journey, it was filled with unforgettable moments. I am proud of the countless sleepless nights and hard work poured into projects, extracurricular activities, and significant achievements over the last four years. 💫

Projects

Color-Based Robot Sorting Powered by CLIP

🎨 Interpret natural language commands to pick, place, and sort objects by color!
🤖 This system leverages CLIP for visual-language understanding, enabling a robot to respond to human instructions and sort objects by color in real time.
⚡ Combines computer vision and natural language processing for flexible, intuitive robot teleoperation.

Python CLIP Computer Vision Manipulators Natural Language Processing

View Code Watch Demo

Dual-Arm Hand Tracking Control System

🖐️ Control two robotic arms in real time using your hands!
🤖 Combines MediaPipe hand tracking with MoveIt Servo for smooth, intuitive dual-arm panda robot control.
⚡ Enables natural, real-time teleoperation of both arms for advanced manipulation tasks.

Python ROS MoveIt MediaPipe Real-Time Control Computer Vision Manipulators

View Code Watch Demo

Optimal Control of Chronos Car

🚗 Designed and implemented planners, including a Path Following Planner for autonomous navigation with MoCap marker tracking.
🧠 Applied Reinforcement Learning (DQN and DDPG) to train a 1/28 scale car to navigate tracks autonomously using Stable-Baselines3.
🌟 Enhanced the simulation environment to optimize RL training, iterating extensively with different models.
🎯 Manipulated the reward function to achieve significant performance gains.

Python RL MoCap Path Planning

Watch Demo View Presentation

NMPC Control of 2-DoF Robot

🛠️ Implemented a Nonlinear Model Predictive Controller (NMPC) using CasADi in MATLAB.
🎯 Solved the problem under varying conditions: minimal energy consumption, minimal time, and model uncertainties with external noise.
📚 Project completed as part of the Nonlinear MPC course, showcasing advanced control techniques.

MATLAB CasADi NMPC Control

View Code View Detailed Info

Instance Segmentation for Waste Detection in the Wild

🧠 Developed Mask-RCNN model with Detectron2 library for waste detection in the wild.
🌟 Addressed challenges: diverse objects, cluttered backgrounds, and varying lighting.
☁️ Deployed on the cloud using Streamlit for accessibility and scalability.

Python PyTorch Computer Vision Deep Learning Instance Segmentation Data Augmentation

View Code View Report View Presentation

Autonomous Mobile Robot for Slug Detection and Collection

🤖 Designed and implemented a ROS-based robot for detecting and collecting slugs to prevent agricultural damage.
📸 Integrated YOLOv5 for real-time slug detection with high accuracy.
🚜 Engineered a ramp and roller mechanism for safe collection, ensuring slug safety.
🌍 Equipped with a GPS-based navigation system to navigate rough agricultural terrains.
🌱 Prioritized sustainability with a scalable, cost-effective alternative to conventional pest control methods.

ROS YOLOv5 Python Navigation Teamwork & Collaboration Mapping & Localization Path Planning Embedded Systems Object Detection

View Code Watch Demo Read Paper View Presentation

Robot Pose Estimation Research

🔬 Conducted in-depth research on robot pose estimation using external camera systems during a master's course.
📚 Comprehensive literature review of three research papers:
📄 "Camera-to-Robot Pose Estimation from a Single Image" Timothy E. Lee et al. | NVIDIA + CMU | Dec 2019
📄 "Markerless Camera-to-Robot Pose Estimation via Self-supervised Sim-to-Real Transfer" Jingpei Lu et al. | UC San Diego | Mar 2023
📄 "External Camera-based Mobile Robot Pose Estimation for Collaborative Perception" Simon Bultmann et al. | Uni Bonn | ICRA June 2023

Computer Vision Pose Estimation Research Camera Tracking Robotics

View Presentation

Publications

Design and Implementation of an Autonomous Mobile Robot for Slug Detection and Safe Collection to Prevent Agricultural Damage
Journal: International Journal of Mechanical Engineering and Robotics Research (IJMERR), Presented at IEEE ICMRE 2024, Milan, Italy.

Date: February 28, 2024

This research presents a novel, eco-friendly robotic solution to address agricultural slug damage. The robot, developed using Robot Operating System (ROS) and integrated with YOLOv5, autonomously detects and collects slugs without harming them. It employs a ramp and roller mechanism for safe collection and features a GPS-based navigation system designed for rough agricultural terrains. This innovative approach prioritizes sustainability and slug safety, offering a scalable and cost-effective alternative to conventional pest control methods.

ROS YOLOv5 Python Navigation Teamwork & Collaboration Mapping & Localization Path Planning Embedded Systems Object Detection

Read Publication View Presentation View Demo

A New Approach for Seven-Segment Display Control Based on One Control Pin
Conference: 2021 IEEE 3rd International Multidisciplinary Conference on Engineering Technology (IMCET)

Date: January 10, 2022

This paper introduces a novel method to simplify the control of seven-segment displays by reducing the number of required control pins to just one. The proposed approach eliminates the need for complex communication protocols by utilizing an RC filter, combinational logic circuit, flash converter, and a BCD-to-seven-segment encoder. The implementation, tested using an Arduino Uno, demonstrates significant reductions in hardware complexity and pin usage, paving the way for more efficient integration of seven-segment displays in various applications.

Arduino Combinational Logic Seven-Segment Display Hardware Simplification Microcontroller Programming Digital Logic Circuits Embedded Systems Problem-Solving

Read Publication

Nabil Miri

Robotics Software Engineer

About Me

Technical Skills

Programming Languages

Robotics & Control

Machine Learning & Vision

Cloud & DevOps

Experience

Fraunhofer IPA - Master Thesis

Nov 2024 – Apr 2025

Stuttgart, Germany

Fraunhofer IPA - M.Sc. Thesis

📝 Project Title:

🔍 Problem Statement:

⚙️ Methodology:

🛠️ Implementation:

📊 Results:

🤖 Platform:

🧪 Experiments

🔗 Links

BMW AG - Robotics Intern

Apr 2024 - Sep 2024

Munich, Germany

BMW Group - Robotics Intern ⚙️

🌟 Summary:

📄 Key Takeaways:

🛠️ Technical Highlights:

🧩 Collaborative Tasks:

TU Dortmund - Research Assistant

Apr 2023 – Mar 2024

Dortmund, Germany

Student Research Assistant 📚

🌟 Summary:

🛠️ Technical Highlights:

🛠️ Skills & Tools:

Research Internship

May 2022 – Oct 2022

Troyes, France

Research Internship [Remote]

Jun 2022 – Sep 2022

Remote

Education

Oct 2022 - Apr 2024

Dortmund, Germany

M.Sc. Automation and Robotics

🛠️ Activities & Achievements

🏆 Academic Grade

📚 Key Courses

🔧 Skills Gained

Aug. 2022 - Nov. 2022

Remote

Professional Certificate - Machine Learning

2018 – 2022

Beirut, Lebanon

B.Sc. Mechatronics, Robotics, and Automation Engineering

🎓 Bachelor's in Mechatronics, Robotics, and Automation Engineering

🌟 Final Year Project (FYP)

🏆 Key Achievements

📜 Certifications

🤝 Leadership Role

📄 Research Papers

💬 Reflections

Projects

Color-Based Robot Sorting Powered by CLIP

Dual-Arm Hand Tracking Control System

Optimal Control of Chronos Car

NMPC Control of 2-DoF Robot

Instance Segmentation for Waste Detection in the Wild

Autonomous Mobile Robot for Slug Detection and Collection

Robot Pose Estimation Research

Publications

Certificates & Awards

🥉 3rd Place in TUD Startup Weekend 2023

🥇 1st Place in Beirut AI & Microsoft AI Bootcamp

🥈 2nd Place in Beirut AI University Hackathon 2020

RHU President's Honor List

Get in Touch

Let's Connect!