How to Choose and Plan Your Final Year Project: Proposal to Viva
A practical guide to choosing a final year project idea, writing the proposal, planning a timeline and surviving the viva. Check your department's format.
On this page
Save this guide before your group sits down to pick a title. Most final year projects do not fail because the students are weak. They fail because the idea was too big, the parts arrived late, the team never agreed who does what, or the demo was practised once. All four problems can be fixed with planning in the first two weeks.
This guide is general advice for engineering and IT students, including those in Nepal. It does not describe any one university's rules. Deadlines, report format, team size and marking differ between universities and departments, so check your department's format and ask your supervisor early. What follows is the practical part that works almost everywhere.
Step 1: Choose the idea
Start with a problem, not a gadget
"Build an IoT thing" is not a project. "Farmers lose water because tanks overflow" is a problem. A good problem has a user, a pain and a way to measure improvement.
Ideas can come from:
- Your own home, hostel or college (water, power cuts, parking, attendance).
- Local sectors such as agriculture, hydropower, tourism, traffic and earthquake safety.
- Your teachers' or seniors' unfinished work that can be extended.
- A skill you want on your CV, such as embedded firmware or machine learning.
Filter with five questions
Scroll sideways to see all columns.
| Question | Good sign | Warning sign |
|---|---|---|
| Can we explain it in two sentences? | Yes | Needs a five-minute story |
| Do we understand each block? | Yes, or can learn it in a month | Several blocks are magic |
| Can we get the parts or data in time? | Yes, within two weeks | Needs imported or rare parts |
| Can we demo it in five minutes? | Yes | Needs special conditions |
| Is there room for our own contribution? | Yes | Just copies a known project |
If three or more land in the warning column, change the idea now.
Size it correctly
Think of three levels:
- Minimum: the smallest version that still works and proves the idea.
- Target: minimum plus one or two features.
- Dream: the extras you only do if everything else is done.
Promise the target in your proposal, build the minimum first, and treat the dream as a bonus.
Step 2: Write the proposal
Formats vary, but most proposals cover similar ground. Use your department's template if one exists.
Scroll sideways to see all columns.
| Section | What to write |
|---|---|
| Title | Short and specific |
| Background | Why the problem matters, in a few lines |
| Problem statement | One clear paragraph |
| Objectives | Three or four measurable goals |
| Scope and limitations | What you will not do |
| Methodology | Block diagram, hardware, software, algorithm |
| Expected outcome | What the system can do and how you will test it |
| Timeline | A week-by-week or month-by-month plan |
| Budget and parts | A list with quantities, no guessing at the last minute |
| References | Papers, datasheets, books and websites you actually used |
Write measurable objectives
Weak: "To make a smart system for farmers." Better: "To measure soil moisture every 10 minutes and switch a pump on below a set threshold, and to compare water use with manual watering over two weeks."
If you cannot test an objective, rewrite it.
Step 3: Plan the timeline
Do not divide time equally. Hardware and integration take longer than people expect.
Scroll sideways to see all columns.
| Phase | Rough share of time | Output |
|---|---|---|
| Research and design | 15% | Block diagram, parts list, approved proposal |
| Component testing | 15% | Each sensor and module works alone |
| Integration | 25% | Everything works together |
| Testing and improvement | 20% | Measurements, bug fixes |
| Documentation | 15% | Report, diagrams, screenshots |
| Rehearsal and buffer | 10% | Demo practice, spare time |
Add buffer. Parts arrive late, boards burn out, and a team member will fall sick before the deadline. Expect at least one of the three.
Use simple tools
- A shared spreadsheet or Kanban board with "To do, Doing, Done".
- A Git repository for code with regular commits, so everyone can see who built what.
- A weekly 20-minute team check-in with a fixed time.
Step 4: Split the work
Avoid the "one hero, three passengers" pattern. Examiners often ask each member a question on any part of the system.
- Give each person a main area (hardware, firmware, software or dashboard, report) and a backup area.
- Every member should be able to explain the full block diagram.
- Rotate who runs the demo during rehearsals.
Step 5: Build smart
- Test each module on a breadboard alone.
- Add one module at a time to the working system.
- Save a working version of the code before every big change.
- Keep a test log with date, condition, expected result and actual result.
- Move from breadboard to a soldered board or neat enclosure well before the demo. Loose jumper wires fail at the worst moment.
Hardware safety
Demonstrate at low voltage wherever possible. Only a qualified electrician should handle mains wiring. Get qualified help with large batteries and high-current motors. Use fuses, correct wire gauge and insulated enclosures.
Step 6: Documentation
Start the report in week one, not week ten. Write a page every week while the details are fresh.
Keep a folder with:
- Final block diagram and circuit diagram
- Flowcharts of main code logic
- Photos of build stages
- Test result tables and graphs
- A list of references
- Source code with comments
Cite sources honestly. Code or text taken from elsewhere should be listed, and your report should be written in your own words. Check your institution's rule on plagiarism.
Step 7: Prepare for the viva
Questions you should expect
- Why did you choose this project? What problem does it solve?
- Explain the block diagram, in order.
- Why this sensor or microcontroller, and what is one alternative?
- What are the limits and sources of error?
- What would you improve with more time?
- What did each team member do?
- Show a test result. How did you measure it?
Viva habits that help
- Practise out loud. Rehearse the five-minute explanation in front of juniors.
- Know the theory behind the formulas you used, such as Ohm's law, sampling or the algorithm.
- Carry a backup: spare cables, a power bank, a recorded demo video and printed diagrams.
- Say "I don't know, but I would check the datasheet" if stuck. Honesty works better than guessing.
- Do not hide limits. Examiners respect a team that says "this sensor drifts in humidity, so we calibrated it".
Common mistakes
Scroll sideways to see all columns.
| Mistake | Fix |
|---|---|
| Choosing the title before the problem | Define the problem first |
| Buying all parts on day one | Buy for the minimum version first |
| Leaving the report to the end | Write weekly |
| Demo depends on the college Wi-Fi | Use a phone hotspot or run offline |
| No test data | Log measurements from week four |
| One member knows everything | Rotate explanations |
FAQ
How early should I start?
As early as your department allows. Even a rough idea and a parts list a few weeks ahead of the official start gives you a major advantage.
Should I pick a hardware or software project?
Pick what matches your interests and your department's expectations. Hardware gives a strong live demo but has more risk with parts and wiring. Software is safer on logistics but needs a clear algorithm or technical depth.
What if my project stops working before the viva?
Keep a backup: a recorded video of the working system, saved test results and photos. Explain calmly what happened and show the evidence.
How do I know the report format?
Check your department's format. Ask your supervisor or seniors for the latest template and follow it exactly.