Portfolio · Industry

Engineering intelligent AI solutions.

Mission briefs from drone autonomy, edge inference and satellite Earth observation.

MSN-01 · Drone · Edge AI

Embedded AI human detection

Real-time drone-based human detection system

Platform
Jetson Xavier NX
Airframe
HolyBro
Runtime
TensorRT
  • Computer vision
  • Embedded systems
  • TensorRT
  • Drone integration

Overview

To create an autonomous drone-based surveillance system capable of real-time human detection, I engineered a complete end-to-end solution: assembling a HolyBro drone, designing and 3D-printing a custom payload module for NVIDIA Jetson Xavier NX integration, and implementing TensorRT-optimised AI models for efficient human detection during flight.

Achievements

  • Hardware integration. Assembled the HolyBro drone with a custom 3D-printed payload module for seamless Jetson integration.
  • Real-time performance. Optimised inference speeds with TensorRT on resource-constrained embedded hardware.
  • AI model deployment. Accurate human detection from an aerial perspective during live flight operations.

MSN-02 · Drone · Mission control

Kiosk ground station software

One-click autonomous drone mission control

Framework
Qt
Protocol
MAVLink
Mode
Kiosk UI
  • Qt development
  • UI/UX design
  • MAVLink
  • Autonomous systems

Overview

To streamline drone operations and reduce operator complexity, I built a complete ground station that bridges sophisticated autonomous flight and user-friendly operation: a Qt desktop application with custom UI components, MAVLink integration for real-time telemetry and control, and a workflow that executes pre-planned autonomous missions from a single click.

Achievements

  • One-click operation. Complex autonomous mission execution simplified to a single click for non-technical operators.
  • MAVLink integration. Full MAVLink protocol stack for reliable drone communication and telemetry monitoring.
  • Custom Qt interface. Kiosk-style UI with real-time status, mission planning tools and emergency control.

MSN-03 · Satellite · Geospatial

AI-powered change detection

Satellite imagery analysis and monitoring

Input
Multi-temporal EO
Task
Change detection
Scale
Large datasets
  • Deep learning
  • Model optimisation
  • Computer vision
  • Satellite imagery

Overview

To address the growing need for automated monitoring of environmental and urban change, I developed an AI change detection system that analyses satellite imagery over time: custom deep learning architectures optimised for change detection, multi-temporal image analysis, and robust preprocessing pipelines for diverse imagery formats and resolutions.

Achievements

  • Enhanced accuracy. Custom models significantly improved change detection accuracy over traditional methods.
  • Model optimisation. Compression and optimisation strategies for efficient deployment without losing performance.
  • Scalable solution. Algorithms that process large-scale satellite imagery datasets for real-world monitoring.

// In the pipeline

More projects coming soon.

Showcase material in preparation for 3D reconstruction, autonomous navigation, agricultural AI, smart city monitoring and mobile AI.

  1. Q.013D Reconstruction Systems (NVIDIA Minkowski Engine)Pending
  2. Q.02Autonomous Wheelchair Navigation (Calibration)Pending
  3. Q.03Agricultural AI — Radish Wilt DetectionPending
  4. Q.04Recycling Classification SystemPending
  5. Q.05Smart City Video ProcessingPending
  6. Q.06Sewer Inspection AIPending
  7. Q.07Cross-Platform Mobile AIPending
  8. Q.08Thermal Image ProcessingPending
  9. Q.09Tablet Inspection SystemsPending

// Open channel

Let's put intelligence in orbit — and in the air.

Launching an AI product, optimising an existing system or exploring research? Let's build something intelligent, scalable and impactful together.