HAECO CEO presenting the Outstanding Innovation Certificate
Grand Prize Winner

AWS AI Hackathon Hong Kong 2025

HAECO Bay Management System

Built in 14 days โ€ข 130+ shortlisted teams โ€ข Hong Kong's first and largest AWS AI Hackathon

Project Snapshot
๐Ÿ† Grand Prize 130+ Teams 14 Days 4 Interviews

The Challenge

Aircraft Bay Assignment at HAECO

HAECO operates in a large-scale aircraft base maintenance environment, delivering maintenance services under complex operational, regulatory, and resource constraints. Effective bay assignment requires balancing aircraft compatibility, task urgency, and dynamic operational changes.

โš ๏ธ

Manual Planning Complexity

Bay assignment is done manually with time pressures, leading to errors and potential regulation breaches

๐Ÿ”ง

Multiple Constraints

Each bay has constraints on aircraft models and maintenance task types, plus geographical dependencies

๐Ÿ“Š

Dynamic Operations

Frequent task reprioritization due to unpredictable external factors creates planning bottlenecks

Our Solution

๐Ÿค–

AI-Powered Optimization

Automated bay assignment using AI that considers work type compatibility, urgency, bay dependencies, and operational factors

๐Ÿ“Š

Real-Time Visualization

Interactive dashboard showing bay utilization, aircraft schedules, and system statistics at a glance

โšก

Zero Installation

Web-based system that opens directly in browser - fast, secure, and accessible anywhere

๐Ÿ’ฌ

AI Assistant

Built-in chatbot for instant help, conflict detection, and operational insights

3-Minute Demo Video

Key Features

1

Dashboard

Zero-install web app opens directly in browser. Planners instantly see upcoming aircraft, bay utilization, and system statisticsโ€”all in one place.

2

Data Import

Download CSV template, fill in aircraft registration and task details, then import. System validates and processes everything locallyโ€”fast and secure.

3

Schedule & Optimization

Bay assignments generated automatically. System considers aircraft model constraints, bay dependencies, and task urgency. Fewer errors, reduced unnecessary movements, operational efficiency insights.

  • Consistent and stable assignment outcomes during testing
  • Efficiency and cost optimization
  • Auto-assign or AI-powered optimizer
4

Bay Status & Layout

Real-time bay availability and aircraft assignments. Digital replica of HAECO hangar layout with daily view.

  • Bay Utilization Timeline (7-day rolling window)
  • Dynamic layout updates
  • Easy adjustments for delays or early departures
5

Analytics & Reports

Clear insights into bay utilization, cost reports, and operational trends. Managers make confident, data-driven decisions.

  • Bay Type Utilization charts
  • Work Type Distribution
  • Optimization Score tracking
  • Recent Activity Log
6

AI Assistant

Built-in chatbot responds instantly to questions like: "How do I assign an aircraft to a bay?" "What are today's conflicts?" "Show me optimization statistics"

Technology Stack

Frontend

HTML5 CSS3 JavaScript

AI Development

AWS Q Developer Kiro

Features

AI Optimization Real-time Viz Analytics

14-Day Sprint

Requirements & Design

Used Kiro to document system requirements through vibe coding. Validated and clarified specifications to ensure robust, logically consistent design.

Development

Built web-based solution with HTML, CSS, JavaScript. Leveraged Kiro's support to streamline development and focus on core logic and UX.

AI Integration

Integrated AWS Q Developer for intelligent bay assignment. Implemented optimization engine with comprehensive rule-based planning.

Demo & Presentation

Created 3-minute demo video, prepared presentation materials, and delivered system walkthrough with UI design showcase.

Team HAECO

Jason Au-Yeung

Jason Au-Yeung

Derek Szeto

Derek Szeto

Peter Yuen

Peter Yuen

Sahil Kunder

Sahil Kunder

Developer & Demo Presenter

Jason Au-Yeung

Led the technical direction of the hackathon build, connecting solution design, implementation, and the final live presentation into one coherent story.

Key Contributions
  • Solution architecture & core development
  • System integration & live demo presentation
Team Context

Transformation & Technology
Technology Innovation Team

Developer โ€“ Testing & Documentation

Derek Szeto

Strengthened the build through validation, debugging, and documentation support so the prototype stayed stable, explainable, and presentation-ready.

Key Contributions
  • Debugging & testing for system stability
  • Technical documentation & QA support
Team Context

Transformation & Technology
Technology Innovation Team

Business Analyst

Peter Yuen

Grounded the project in business requirements and workflow logic, helping the team shape a solution that matched real operational use cases.

Key Contributions
  • Business requirements & use-case definition
  • Advisory on workflow & solution feasibility
Team Context

Transformation & Technology

Subject Matter Expert (SME)

Sahil Kunder

Brought real maintenance knowledge into the build so the concept stayed connected to actual bay operations, data realities, and on-the-ground constraints.

Key Contributions
  • Operational expertise & maintenance insights
  • Data source & real-world constraints
Team Context

Base Maintenance

Jason Au-Yeung

Developer & Demo Presenter

Derek Szeto

Developer โ€“ Testing & Documentation

Peter Yuen

Business Analyst

Sahil Kunder

Subject Matter Expert (SME)

Jason Au-Yeung

Developer & Demo Presenter

Derek Szeto

Developer โ€“ Testing & Documentation

Peter Yuen

Business Analyst

Sahil Kunder

Subject Matter Expert (SME)

Hackathon Details

AWS AI Hackathon Hong Kong 2025

"Build Tomorrow's Hong Kong with AI"

Official Website aws-hong-kong-hackathon.devpost.com

Timeline

  • Briefing: August 26, 2025
  • Submission: Aug 15 - Sep 30, 2025
  • Judging: Oct 2 - Oct 14, 2025
  • Winners: October 16, 2025

Requirements

  • Use Amazon Q Developer or Kiro
  • Choose theme or freestyle
  • 3-minute demo video
  • Source code repository

Judging Criteria

  • Business Value & Use Case
  • Implementation Quality
  • Innovation & Originality

Statistics

  • 840 total participants
  • 130+ shortlisted teams
  • Hong Kong's first AWS AI Hackathon
  • 4 non-cash prizes

Impact & Recognition

๐ŸŽ™๏ธ

4 Media Interviews

  • unwire.hk - Tech media coverage
  • SCMP - South China Morning Post
  • HAECO - HiHAECO Post, Lean Day
  • AWS Hong Kong - Official interview
๐ŸŒ

Global Exposure

Represented HAECO at AWS re:Invent (Las Vegas), interviewed on-site and featured in unwire.hk covering the HAECO Bay Management System.

Demonstrated ability to explain complex operational systems to a global audience.

โ–ถ๏ธ YouTube Interview
๐Ÿ’ผ

Professional Capability

  • Owned end-to-end solution design under a 14-day hackathon timeline
  • Translated real operational constraints into a working digital prototype
  • Balanced technical feasibility with business requirements
  • Delivered a system suitable for enterprise-level discussion
๐Ÿš€

From Hackathon to Enterprise Thinking

The prototype outlines a clear digitalization direction, demonstrating how data-driven approaches and system design principles can inform future workflow innovation.

Reflection

A short reflection on what this project taught me and how it shaped the way I approach building.

What This Experience Meant

This hackathon was more than a competition for me. It was the first time I experienced how quickly an idea could move from concept to a working prototype when the goal was clear and the team was fully committed. Winning the project gave me confidence, but more importantly, it showed me that I enjoy building solutions that respond to real operational needs.

What Challenged Me

The biggest challenge was working under a very short timeline while still trying to make the project practical, presentable, and meaningful. It was not only about building a system, but also about explaining the logic clearly, shaping the story of the project, and making the final output understandable to others.

What I Will Carry Forward

After this experience, I want to keep developing projects that combine operations, data, and AI in a practical way. I also want to keep improving how I communicate ideas, present systems, and turn technical work into something that creates visible value for real users.