The CSE ideas span eight domains, the AI and machine learning ideas are tagged by level, the IoT builds name the hardware, and the ECE ideas cover VLSI, embedded, communication, DSP and RF. Every idea says what you would build it with, such as a public dataset, an API or specific sensors, so you can judge the scope before you take it to your guide. They are topic ideas, not projects we have delivered.
Which branch list should I use?
The four lists overlap a little in AI and embedded systems, but each is written for a different decision. Pick the one that matches your branch and what you are building. ECE, EnTC and EXTC students have their own list for VLSI, communication, DSP and RF; EEE students building with a board and sensors will find the IoT list the closest fit.
| List | Ideas | Best for | Each idea gives you |
|---|---|---|---|
| CSE and IT | 53 | B.Tech and B.E. computer science and IT teams still choosing a domain | A one-line pitch, a difficulty level, a typical stack, a public dataset or API and a viva angle |
| AI and ML | 49 | B.Tech mini and major projects and M.Tech work in AI | A level, a named dataset, the metric to report and a research extension |
| IoT | 33 | Final year teams building with a board and sensors, including ECE and EEE | The board, named sensors, a dashboard, a difficulty level and a rough cost band |
| ECE and EnTC | 27 | B.Tech and B.E. ECE, EnTC and EXTC teams who want to build and measure | The board or tool, what you measure, a viva angle, a level and a rough cost band where parts are bought |
IoT and ECE cost bands are rough totals for the electronics only, not quotes. Check current prices before you buy.
Past the final year? Our PhD topics in computer science are written as research questions, each with recent surveys, a public dataset and a feasibility note, to discuss with a supervisor.
Final year project ideas for engineering students: three picks from each list
A quick way in if you are not sure where to start. These twelve were picked for a spread of levels and domains, not ranked; each links to its section, where you will find the full idea with the data, stack or sensors and a viva angle.
| Idea | List | Level | Built with |
|---|---|---|---|
| Clash-free timetable generator | CSE and IT | Intermediate | Google OR-Tools (CP-SAT), Django, React, your department’s real constraints |
| Pothole and crack detector for road videos | CSE and IT | Intermediate | YOLO and OpenCV, trained on RDD2022, which includes road images from India |
| Q&A assistant for India’s new criminal laws | CSE and IT | Advanced | Retrieval-augmented generation with an open model, over bare acts from India Code |
| Plant disease detection that works in the field | AI and ML | B.Tech major | Train on PlantVillage, test on PlantDoc field photos, report macro-F1 |
| Hindi-English code-mixed sentiment | AI and ML | B.Tech major | SemEval-2020 Task 9 (Hinglish), MuRIL or IndicBERT against XLM-R |
| Drift monitoring with automatic retraining | AI and ML | B.Tech major | NYC taxi trip records, a drift score and MLflow tracking |
| Water-tank level monitor and pump control | IoT | Intermediate, low cost | ESP32, a waterproof ultrasonic sensor, a flow sensor, a relay and Blynk |
| Drip irrigation that checks the forecast | IoT | Beginner, low cost | ESP32, a capacitive soil-moisture sensor, DHT22, a 12 V solenoid valve and the Open-Meteo API |
| Fall detector for elderly people | IoT | Intermediate, low cost | ESP32, an MPU6050 accelerometer and gyroscope, a buzzer and Telegram or Blynk alerts |
| Fixed-point FIR filter on an FPGA | ECE and EnTC | Intermediate, free in simulation | Coefficients from MATLAB or SciPy, Verilog or VHDL checked against the software model, the lab FPGA board |
| LoRa link budget checked in the field | ECE and EnTC | Intermediate, low to medium cost | Two ESP32 boards with SX1276 modules on India’s 865–867 MHz band, RSSI and SNR logged against distance |
| ECG denoising and QRS detection | ECE and EnTC | Intermediate, free | MATLAB or Python with the open MIT-BIH Arrhythmia and Noise Stress Test databases on PhysioNet |
All twelve are topic ideas, not projects we have delivered. IoT and ECE cost bands are for the electronics only; check current prices before you buy.
How do I choose a final year project idea?
Shortlist three ideas, then run each through these checks. The one that passes all four is usually the right one.
Check the data or hardware first
Open the dataset or API link, or price the parts, before you commit. An idea without data you can use is not a project yet.
Size it to your time
A mini project proves one technique works; a major project wraps it in a complete, tested system. Most ideas can be scoped either way.
Name the one result you will show
A metric against a baseline, a live dashboard reading or a working demo. If you cannot say it in a sentence, the scope is still too loose.
Get your guide’s approval in writing
Take a one-page abstract. For M.Tech and IEEE projects, anchor it on a published paper: see how to select a base paper.
How big should a final year project be?
Big enough to show one complete system and one measured result, and small enough to finish, test and explain before your final review. Your syllabus sets the rules on team size, reviews and report length, so read it first; within those rules, this is a sensible scope.
- A mini project proves one technique works: one model on one dataset, or one sensor reading shown on one dashboard, with a short report.
- A major project wraps that technique in a system someone could use: data in, a result out, a simple interface, testing against a baseline, a full report, a PPT and a live demo.
- Every team member should own a part they can explain in the viva, such as the data pipeline, the model, the app or the hardware.
- Leave time at the end. The report, the PPT and demo rehearsals usually take longer than teams expect.
Our final year project ideas for CSE page compares mini and major project scope in more detail.
Are these ideas the same as your delivered projects?
No. The lists are topic ideas to help you choose; they are not projects we have already built, and no idea comes with results. Our delivered work is in eight case studies, all M.Tech and M.E. projects, such as network traffic prediction with a Bi-LSTM and skin disease detection in the browser.
If you want an idea built for your course, with code, report, PPT and viva preparation, see final year projects for CSE and IT students, M.Tech projects or IEEE projects. For writing formats, our research writing guides cover IEEE papers, synopses and theses.
Examiners ask why you chose a topic and what you changed. Add a local dataset, a different metric or one extension of your own, and you will have an answer.
