The Three-Lock Journey

Define. Design. Prove.

Three progressive submission locks, each with a published rubric and a criterion-level evaluation report. The locks never eliminate — they make every team stronger before the next stage.

Start the Journey

A student pinning field photographs and notes into a problem map on a wall.

Two builders debating over a whiteboard system sketch, marker in hand.

A working prototype held up on a phone, its screen glowing, teammates’ proud reflection behind.

Lock 1 · Define

Clarify the problem worth solving.

Everything that follows stands on this foundation.

You submit

Problem statement with evidence
Target users and their current workaround
SDG alignment with the specific target
The AI opportunity, precisely claimed
Expected impact

The rubric scores

Problem clarity (20)
Evidence and user understanding (15)
SDG relevance (15)
Meaningful use of AI (20)
Originality (10) · Feasibility (10) · Potential impact (10)

Lock 2 · Design

Blueprint the complete solution.

Design decisions, not wish lists — including what you will not build.

You submit

Solution blueprint and user journey
Technical architecture diagram
AI approach and data plan
Responsible-AI and privacy plan
Impact metrics for Lock 3

The rubric scores

Improvement from Lock 1 (15)
Solution–problem fit (20)
Technical architecture (20)
Appropriate use of AI (15)
Execution feasibility (15) · Responsible AI (10) · Impact plan (5)

Lock 3 · Prove

Show that it actually works.

The heaviest-weighted stage: half of your final selection score.

You submit

Working prototype with live link
GitHub repository
Demo video and screenshots
Final architecture
Contribution declaration per member
Known limitations, honestly

The rubric scores

Working functionality (25)
AI implementation (20)
Problem–solution fit (15)
User experience (10) · Technical quality (10)
Innovation (10) · SDG impact (5) · Demo clarity (5)

After Lock 3

How the journey ends.

Weights

The selection formula

Lock 1 × 15% + Lock 2 × 20% + Lock 3 × 50% + human review × 15%. Published, fixed, identical in both editions.

Verify

Human verification

People verify AI recommendations before any team is selected. Significant score variance triggers moderation automatically.

Finals

Physical-final selection

Each edition selects its own finalists, who demo in person before that city’s jury.

Prepare for Each Lock

The guides that map to the rubrics.

Lock 1

Problem Definition guide

Finding, narrowing and evidencing a problem worth building for.

Lock 2

Solution Blueprint guide

Turning a validated problem into a buildable design.

Lock 3

Prototype Development guide

Shipping a genuinely working prototype in the window.

Three locks stand between you and the finals stage.

Every one of them makes your project better. Begin with Define.