Design, build and ship one complete mobile app.
In teams of up to 4, you build a mobile app from your own new idea: analysis, design, implementation, testing and a live demo. It is worth 60% of the module.
- Milestones
- 05
- Required features
- 05
- Not counted
- 01
- Deliverables
- 05
Overview
In teams of up to 4, you build a mobile app from your own new idea: analysis, design, implementation, testing and a live demo. It is worth 60% of the module.
The problem
Choose a realistic problem for a specific group of users and get it approved by the instructor. It must be a new idea, not a previous project. A list of ideas will be provided if you need inspiration.
- Architect the solution and communicate it clearly, to people and to AI
- Critically evaluate AI-generated code
- Integrate and refine that code into one coherent, working system
- Present and defend your engineering decisions
Your proposal result
Type your student ID to see your group, the idea approved for it and your feedback. Only your own group is shown.
Submissions
Milestones
- M1
Team + idea
Form a team and submit your idea for approval.
Week 2
- M2
Phase 1
Requirements, diagrams, architecture, data plan, 9 screens.
Week 6
- M3
Build
Storage, Google API, device feature, validation.
Weeks 7–11
- M4
Phase 2
Final app, 3 unit tests, AI engineering log.
Week 12
- M5
Demo
30-minute team demo + individual questions.
Final week
Requirements
- Team of max 4 students, with a new, instructor-approved idea
- Git & GitHub are mandatory: one branch per member, with commits from each member's own account
- At least 9 screens (3 per member), excluding sign-in and registration
- Architecture / design pattern applied. Monolithic = 0 for Phase 2
- Persistent data storage (local and/or cloud)
- One Google API + one device hardware/software feature
- 3 unit tests + a full AI engineering log
Features
Required
5+ functional requirements
Your core features. Login and sign-up do not count.
Google API
Maps, YouTube Data, Analytics, Cloud Storage, Drive, Calendar, Docs, Translate or Forms.
Device feature
Camera, device info, location, sensors, microphone/speakers, a platform-specific feature or inter-app communication.
Persistent storage
File storage
Validation & error handling
Validate every form. Catch failures and show clear messages.
Allowed · not counted
Sign-in / registration
Allowed, but not counted toward screens or requirements.
Technical requirements
- React Native + Expo, TypeScript (as taught in the labs)
- Folder structure reflects your chosen pattern (e.g. MVVM, Provider/Context)
- 3 unit tests: designed, run and reported
- No secrets or API keys committed. Use
.envand.gitignore
UI requirements
- 9+ screens: 3 per member, each correctly implemented
- Consistent styling (StyleSheet or NativeWind)
- Clear feedback for loading, empty and error states
Deliverables
- 01Phase 1 report + UI code (e-learning and printed)
- 02Phase 2 amended report + final code + AI engineering log
- 03Contribution table on page 1 of each report
- 04Signed coursework submission form + Statement of Academic Honesty
- 05Live presentation & demo. Attendance is mandatory
Evaluation
| Criterion | Marks |
|---|---|
| Phase 1 — Analysis, architecture & prototype Use case 5 · Activity 5 · UI 9 · Architecture 6 (+ report). | 30 marks |
| Phase 2 — Implementation, testing & AI log Architecture 5 · Storage 10 · APIs & errors 11 · Tests 9 · AI log 15. | 50 marks |
| Presentation & demo Individual questions on functionality, architecture (group) and AI decisions. Marks can differ between members. | 20 marks |
| Total | 100 marks |
Notes
AI engineering log — the format
For every significant AI interaction that produces code:
| Part | Write down |
|---|---|
| Prompt | Verbatim, with module/feature, constraints and specifics |
| Output | The full code the AI produced |
| Critical analysis | What you changed and why · how you verified it (test, manual check) |
| Rationale | Why you used AI for this part |
Caution
Asking AI to “generate the entire application” without your own modular design is heavily penalised. AI-generated code without its prompt in the log = 0 for the log and −5 from Phase 2 implementation.
Resources
FAQ
How many students per team?
Maximum 4.
Can we use AI?
Yes, as a development aid. Log every significant interaction (prompt → output → your changes → rationale). Reports must be written by you, not AI. Paraphrasing is limited to 20%.
What if I miss the demo?
You are marked absent for the presentation & demo deliverable (20 marks).
Do all members get the same mark?
Not necessarily. Marks depend on individual contribution and discussion, and your Git commits are part of the evidence.