PERSONAL PORTFOLIO

ETHAN SHAHBAZ

Mechatronics engineer.
Founder.

I know a bit about automating systems and building useful things.
SELECTED WORK

Engineering.
In the real world.

Robots that operate independently. Drones that navigate together. Research carried from simulation into physical testing.

01

Autonomous
warehouse robot.

MECHATRONICS
DESIGN / TEST / VALIDATE

A mobile platform.
A system I brought together.

Assembly, ROS integration, autonomous navigation and vision-guided handling, carried through to physical testing.

ROSSLAMA*PyTorch
About the project

I assembled and programmed a Hiwonder JetRover Ultimate platform, then tested its autonomous operation with human supervision removed or kept to a minimum.

The engineering work centred on integrating navigation, perception and manipulation into a working system, with potential applications in warehouses and remote environments.

Platform & hardware
Platform
Hiwonder JetRover Ultimate Kit
Compute
NVIDIA Jetson Orin
Mobility
Mecanum wheels for forward, lateral and rotational movement
Sensing & handling
LiDAR, depth camera and a six-degree-of-freedom robotic arm
Navigation & perception
ROS
Coordination of sensing, navigation and actuation
SLAM
Building a map while estimating the robot’s position
A* planning
Finding a route through the mapped environment
Vision / PyTorch
Object recognition as part of the perception pipeline
Control & validation

The platform’s STM32F407VET6 controller handles low-level motor and servo control, encoder feedback, IMU acquisition and power monitoring.

I worked with wireless manual control and programmed autonomous operation, then tested the integrated system. The animation illustrates the sorting workflow; it is not recorded test footage.

WAREHOUSE / SORT & DISPATCH

Click or tap a box to dispatch. Numbers show queue order.

Organising the shelves0 queued · 0 dispatched
INTERACTIVE CONCEPTFull shelves. Real aisles. Select a parcel for dispatch.
02

Where should a swarm
do its thinking?

ENGINEERING THESIS
SIMULATION / FLIGHT TESTING
AUTONOMOUS SWARM NAVIGATION
Local peer exchangeOffboard telemetry & commands
Shared search

Three drones search together. The visible rover has not yet been identified.

ARCHITECTURE VISUALISATIONConceptual behaviour, not recorded test results
CLOUD vs EDGE vs HYBRID

One research question.
Three architectures.
Three real drones.

My thesis compared edge, cloud and hybrid computing for swarm navigation and collision avoidance, from simulation to physical tests with three drones.

Edge architecture

Select an architecture to explore how information moves through the system.

Research scope

My thesis investigated edge computing strategies for autonomous swarm navigation and collision avoidance. I compared cloud, edge and hybrid architectures in simulation, then tested those approaches with three physical drones.

The central question was where computation should happen when several autonomous vehicles must coordinate.

Three computing architectures
Cloud
Shared state and navigation decisions processed offboard
Edge
Local navigation decisions processed close to each vehicle
Hybrid
Local responses combined with shared offboard coordination
Comparison
Architecture placement, communication and collision avoidance
Reference implementation
Onboard
Raspberry Pi 4 companion computer running ROS 2 / Python
Offboard
A remote Linux compute service for shared planning
Flight control
A dedicated flight controller retains stabilisation
Modelling
MATLAB / Simulink for vehicle modelling and navigation simulation
From simulation to flight

Simulation provided a way to compare the three architectures before physical testing with a three-drone swarm.

The concept here shows shared search coverage, target acquisition and coordinated tracking. Switching architecture changes where planning data and decisions flow.

03

A ground robot.
A wireless connection.

WIRELESS SYSTEMS
CONTROL / COMMUNICATION

Drive the robot.
Understand what it’s doing.

A responsive wireless control interface with sensor information and live telemetry in one view.

TELEOPERATIONSENSORSGUI DESIGN
About the project

I designed a graphical user interface to control a wireless ground vehicle and bring its sensor information together in one place.

The interface combined responsive driving with visibility into velocity, acceleration, temperature, power draw and other telemetry, making the robot’s behaviour easier to understand while operating it.

Control & telemetry
Drive
Responsive directional commands from the operator
Motion
Velocity and acceleration feedback
Health
Temperature, power draw and battery status
Interface
Controls and sensor data visible together in a single GUI
Reference implementation
Controller
ESP32 with a local motor-control loop
Connection
Wi-Fi between the vehicle and operator interface
Messages
Compact movement commands and timestamped telemetry
Behaviour
Telemetry updates independently from driving input
Try the control interface

Drive through the field using the arrow keys or W A S D while the control area is in view. On touchscreens, hold the direction buttons. Recover three field data capsules in any order, then return to the controller and stop to upload. Watch the signal weaken with distance and when buildings interrupt the path.

Press Esc to release keyboard control. All values in this browser demonstration are simulated.

ROVER CONTROL / TELEMETRYSIMULATED LIVE DEMO
FIELD DATA RECOVERYRecover 3 data capsules, then return to the controller
0 / 3 RECOVERED
VEHICLE TELEMETRYSIMULATED
SPEED
0.00m/s
ACCELERATION
0.00m/s²
TEMPERATURE
28.0°C
POWER DRAW
0.0W
BATTERY
100%
LINK STRENGTH
100%

Strong · −38 dBm

RECENT SPEED
Collect the data capsules
KEYBOARD AVAILABLE IN VIEWArrows / W A S D to drive. Esc to release.

WORK & INDEPENDENT VENTURES

Experience.

Engineering, technology
and independent work.

CAPABILITIES

Skills & tools.

Engineering, software
and professional practice.

Engineering

  • Embedded & connected systems

    • MCUs & MPUs
    • Arduino
    • ESP32
    • Raspberry Pi
    • Embedded systems
    • Wireless communications
  • Robotics & integration

    • Robotics
    • Automation
    • PLCs
    • Systems integration
    • Sensor fusion
  • Design & prototyping

    • CAD
    • 3D printing
    • Prototyping
  • Production & quality

    • Cleanroom experience
    • ISO 9001 quality management
    • Root-cause methodologies

Software & data

  • Programming

    • Python
    • Java
    • C/C++
    • Object-oriented programming
    • Data structures & algorithms
  • Modelling & robotics

    • MATLAB
    • Simulink
    • ROS 2
    • ABB RobotStudio
    • COMSOL Multiphysics
  • CAD & electronics

    • SolidWorks
    • Altium 365
    • Creo
  • Development platforms

    • GitHub
    • Google Firebase
    • Vercel
    • Supabase
  • AI-assisted development

    • Claude Code
    • OpenAI Codex
    • AI applications
  • Team & business tools

    • Jira
    • Confluence
    • Salesforce
    • Notion
    • Microsoft 365
    • Google Workspace

Professional practice

  • Communication & collaboration

    • Effective communication
    • Cross-team collaboration
    • Stakeholder engagement
    • Technical documentation
    • Presenting technical findings
  • Judgement & organisation

    • Critical thinking
    • Attention to detail
    • Organised & systematic
    • Continuous improvement
  • Leadership & growth

    • Leadership
    • Training & mentoring
    • Adaptability & resilience
    • Eagerness to learn
BACKGROUND

A little
about me.

An engineer by training.
A hands-on learner by nature.

I’m Ethan, a mechatronics engineer and founder. I like understanding how things work, then having a go at building something useful.

That has taken me from warehouse robots and wireless systems to a thesis on autonomous drone swarms. I’m interested in the whole process, including the testing, small improvements and things that don’t work the first time.

I take pride in what I make and how I present it. This portfolio is a collection of that work, with more of the story to come.

BUILDTESTLEARN
Ethan ShahbazENGINEER / ENTREPRENEUR
HAVE AN OPPORTUNITY IN MIND?LET’S START A CONVERSATION

LET’S TALK

Engineering opportunities. Collaborations.
Or a conversation about my work.

LinkedIn