Sherif Fathey

Robotics Software Engineer

I build reliable autonomy pipelines and software stacks for mobile robots: ROS 2 navigation, Behavior Trees, SLAM, multi-sensor fusion, and autonomous docking from simulation to production AMRs.

Frontier Exploration: Autonomous mapping in progress...

Portrait of Sherif Fathey

About

I'm a Robotics Software Engineer with 2+ years of experience specializing in autonomous mobile robots, ROS 2 navigation stacks, and intelligent behavior orchestration.

Currently at Arche Robotics, I develop autonomous AprilTag docking systems using Behavior Trees and SMACH state machines, build custom Nav2 costmap layers with dynamic runtime obstacle zones, and configure multi-sensor fusion (EKF, LiDAR, wheel encoders, IMU) for robust AMR localization and mapping.

Beyond industry systems, I am passionate about advancing the robotics community. I serve as Lead Instructor at the Arabian Robotics and ROS Meetup, delivering deep-dive workshops on ROS 2 and BehaviorTree.CPP, and I am pursuing an M.Sc. in Mechatronics and Robotics at Ain Shams University.

Location
Dubai, UAE · Open to relocation · Working remotely worldwide
Experience
2 years building autonomous robots & ROS 2 stacks
Education
M.Sc. in Mechatronics & Robotics — Ain Shams University (Ongoing)
B.Sc. in Mechatronics Engineering — Sinai University

Experience

  1. to present

    Robotics Software Engineer

    Arche Robotics Full-time, Turkey(remote)

    • Developed an autonomous docking system in ROS 2 using AprilTag detection and Behavior Trees with multi-attempt alignment logic.
    • Architected robot workflows using ROS 2 Behavior Trees for Nav2 navigation, task scheduling, and real-time health monitoring.
    • Built SMACH-based Finite State Machines (FSM) for high-level robot behavior control and mission management.
    • Engineered mapping pipelines with slam_toolbox, AMCL localization, and EKF-based multi-sensor fusion.
    • Extended the Nav2 virtual_costmap_layer with dynamic runtime controls and custom exclusion zones.
    • Integrated Gazebo (GZ Sim) and ros2_control for high-fidelity physical simulation and hardware-in-the-loop testing.
    • Enhanced a Debian-based CLI tool to streamline automated robot configuration and deployment workflows.
  2. to present

    Lead Robotics Instructor & Technical Mentor

    Arabian Robotics and ROS Meetup Voluntary

    • Led comprehensive ROS, ROS 2, and Linux training programs for robotics engineers and students.
    • Designed technical workshops covering ROS Topics, Services, Actions, Messages, and URDF kinematic modeling.
    • Developed simulation environments in Gazebo and mentored engineers in BehaviorTree.CPP.
  3. to

    Graduation Project Mentorship

    VALEO Automotive Mentorship

    • Guided the engineering team in developing an AI-driven Driver Monitoring System using MediaPipe.
    • Achieved ~95% drowsiness detection accuracy and integrated facial recognition for vehicle access control.
    • Deployed and optimized the complete computer vision pipeline on Raspberry Pi 5.

Projects

Featured autonomous systems, open-source ROS 2 packages, and robotics research. Live star counts sync from GitHub.

BTick — Behavior Tree Visualizer & Debugger

BTick is a real-time visualization, debugging, and authoring platform for BehaviorTree.CPP, built for robotics development and ROS 2 applications. It provides an interactive environment for monitoring, debugging, and controlling Behavior Tree execution across complex robotics systems.

Key features
  • Automatic Tree Discovery & Multi-Tree Monitoring: Automatically discover and monitor multiple Behavior Trees and nested subtrees in real time.
  • Interactive Execution Debugging: Visualize execution flow and node statuses, set breakpoints, step through execution, and override node statuses to test different scenarios.
  • Live Runtime Editing: Inspect and modify blackboard variables and supported node input ports during execution.
  • Recording, Replay & Performance Analysis: Record execution sessions, replay transitions through a timeline, and analyze node timing to identify potential bottlenecks.
  • Visual Behavior Tree XML Editor: Create, edit, and validate BehaviorTree.CPP v4 XML files through an interactive visual interface.
  • Multi-Protocol & ROS 2 Integration: Support native BTick communication, Groot2 compatibility, and Nav2 Behavior Tree logging.
  • Cross-Platform & Multi-Device Access: Run BTick through web and desktop applications, with responsive mobile and tablet interfaces, QR-code pairing, and synchronized debugging across devices.
  • ROS 2
  • BehaviorTree.CPP
  • C++
  • React
  • TypeScript
  • Electron
  • WebSocket
  • Nav2

VirtuLayer — Virtual Costmap Zone Editor

A web-based visual editor for drawing, configuring, and managing virtual costmap shapes on ROS occupancy maps. Enables robotics engineers to interactively define restricted zones, keep-out buffers, and dynamic cost regions with instant export to YAML for nav2_virtual_layer.

Key features
  • Interactive In-Browser Drawing: Draw circles, convex polygons, and linear barriers directly on standard ROS .pgm/.yaml map pairs.
  • Fine D-Pad Nudging & Cost Tuning: Precision coordinate nudging in meters, per-shape cost weights (1–254), duration timers, and live shape manager table.
  • Instant Nav2 YAML Import/Export: Export ready-to-load virtual_layer_forms.yaml configurations with full Undo/Redo history stack.
  • Nav2 Costmap
  • ROS 2
  • Web Application
  • JavaScript
  • Canvas API

Nav2 Virtual Layer

A production Nav2 costmap plugin for ROS 2 that creates dynamic virtual obstacles and cost zones using circles, lines, and polygons with runtime configuration support. Eliminates static map re-editing when marking restricted or temporary operational zones.

Key capabilities
  • Dynamic Shape Injection: Inject circular forbidden buffers, linear barriers, and arbitrary convex polygons into local and global costmaps.
  • Runtime Lifecycle Controls: Add, clear, and modify navigation zones on the fly via ROS 2 service calls without node restarts.
  • Full Nav2 Stack Compatibility: Compatible with both ROS 2 Humble and Jazzy navigation stacks with zero costmap inflation overhead.
  • ROS 2 Humble & Jazzy
  • Nav2 Costmap2D
  • C++
  • Pluginlib

SDF2MAP

An automated tool for extracting high-resolution ROS-compatible occupancy grid maps (.pgm and .yaml) directly from Gazebo SDF and World files. Allows robotics engineers to generate perfect ground-truth navigation maps instantly without driving robots through manual SLAM passes.

Key capabilities
  • Direct SDF Geometry Parsing: Parses box, cylinder, mesh, and wall geometries directly from Gazebo world files.
  • Instant Nav2 Readiness: Outputs ready-to-load YAML map descriptors and PGM bitmaps formatted for Nav2 map_server.
  • Multi-Height Slicing: Configurable Z-slice thresholds to accurately capture obstacles at robot sensor eye-level.
  • ROS 2
  • Gazebo Sim
  • Python
  • Occupancy Grid

ElevoRaBot Multi-Floor Delivery Robot

Autonomous mobile delivery robot designed for complex multi-story indoor facilities. Features automated elevator calling, precise cabin ingress/egress navigation, and dynamic multi-level map switching.

Key capabilities
  • Elevator Dispatch Protocol: Behavior Tree driven negotiation with simulated elevator APIs for call, floor arrival, and door clearance.
  • Confined Space Nav2 Tuning: Specialized costmap layers and tight MPC controllers for maneuvering inside elevator cabins.
  • Full Gazebo World Simulation: Multi-floor building simulation environment complete with operating elevator shafts.
  • ROS 2
  • Nav2
  • Behavior Trees
  • Gazebo
GZ Fuel Model Watch Demo LinkedIn Post Multi-Floor Fleet Navigation ROS 2 / C++

Agentic UAVs — Autonomous Aerial Search & Rescue RIOTU Lab

Autonomous aerial surveillance and search & rescue system powered by LLM-driven Agentic AI workflows for crowd safety during Hajj. The UAV conducts autonomous monitoring to detect pilgrims in distress, executes physical rescue kit air-drops, and dispatches real-time GPS alerts.

Key capabilities
  • LLM-Driven Cognitive Agents: Contextual reasoning and scene understanding to reliably distinguish between healthy pilgrims and individuals requiring medical assistance.
  • Autonomous Intervention & Air-Drop: Real-time automated physical deployment of first aid and water kits directly to detected coordinates without manual pilot intervention.
  • Multi-Agent Coordination & Alerts: Instant dispatching of contextual emergency alerts with GPS telemetry and adaptive multi-UAV mission management.
  • Agentic AI
  • LLMs
  • Autonomous UAV
  • Computer Vision
  • Multi-Agent Systems

More on GitHub

  • Multi-AMR fleet simulation and distributed waypoint dispatching in ROS (Noetic) with Gmapping SLAM, AMCL, and move_base in Gazebo.
  • Real-time trajectory telemetry capture, RViz visualization, and dead-reckoning drift benchmarking platform for autonomous mobile robots.
  • High-precision autonomous charging station docking system integrating AprilTag visual fiducial tracking
  • Automated Waste Sorting System YOLOv8-based waste classification using Raspberry Pi and Arduino Mega, with pneumatic actuation for automated sorting of plastic, paper, and glass.
  • Autonomous Mobile Robot (AMR)Differential-drive physical AMR running ROS Noetic on Jetson Nano with LiDAR/IMU EKF fusion.
  • Vision-controlled robot navigation driven by OpenCV and MediaPipe finger tracking.
View all repositories on GitHub

Skills

Autonomy & Robotics

  • ROS 1 & ROS 2, Nav2, MoveIt 2
  • SLAM (slam_toolbox, GMapping)
  • Localization (AMCL, EKF, robot_localization)
  • Sensor Fusion & Path Planning
  • Behavior Trees & SMACH
  • Multi-Map Navigation & Costmap Layers
  • Autonomous Docking (AprilTag, ArUco)

Programming & Software Development

  • C++ & Python 3
  • JavaScript & TypeScript
  • Qt, React & Electron
  • Vite & Tailwind CSS
  • HTML, CSS & WebSocket
  • OpenCV & MediaPipe
  • SQLite, JSON & Bash

Simulation & Hardware

  • Gazebo
  • URDF / Xacro Robot Modeling
  • NVIDIA Jetson & Raspberry Pi
  • Arduino & Industrial PCs
  • LiDARs, IMUs & Encoders
  • RGB-D & Intel RealSense Cameras

Development Tools & Platforms

  • Linux (Ubuntu, Debian)
  • Git & GitHub Actions
  • CMake & Docker
  • Cubic

Have a robot that needs to map, localize or navigate? Let's talk.

I'm available for robotics engineering roles, navigation stack development, and autonomous systems consulting worldwide. Ready to discuss your robotics project or explore collaboration opportunities? Schedule a free 30-minute consultation to discuss your requirements and how I can help bring your ideas to life.