25 Final Year Project Ideas for Electronics & Computer Engineering
25 final year project ideas for BE electronics and computer students in Nepal, grouped by domain, with hardware, difficulty, viva value and a scoring table.
On this page
Save this list. Every year the same few projects appear in the department hall: "smart home", "obstacle robot", "attendance system". Not because they are bad, but because nobody thought past the first idea. This list gives you 25 ideas across six domains, each with a problem worth solving and a twist that makes your demo different from the next group's.
It is written for engineering students at IOE, Pokhara University, Kathmandu University and similar programmes. Your department decides the required format, scope and team size, so treat these as starting points and confirm with your supervisor early. Difficulty ratings assume a team of three working for roughly one semester.
How to read each idea
- Problem: the one-line reason the project exists.
- Core: main hardware and software.
- Level: Easy, Medium or Hard for a student team.
- Viva value: what you can confidently explain to an examiner.
- Twist: the small extra that makes it yours.
Domain 1: IoT and monitoring
1. Water tank level and pump controller. Problem: overflowing roof tanks waste water. Core: ESP32, JSN-SR04T waterproof ultrasonic sensor, relay module (low-voltage demo), dashboard. Level: Easy. Viva value: sensor physics, hysteresis control. Twist: log refill times and detect leaks when the level drops with no usage.
2. Landslide early-warning node. Problem: hill roads in monsoon give little warning. Core: soil-moisture sensor, MPU6050 tilt, rain sensor, LoRa or GSM. Level: Hard. Viva value: sensor fusion, thresholds, low-power design. Twist: a solar-charged node that sleeps between readings.
3. Smart greenhouse monitor. Problem: tomato and cucumber tunnels swing hot and cold. Core: DHT22, soil moisture, BH1750 light sensor, ESP32. Level: Medium. Viva value: closed-loop control. Twist: automatic side-curtain motor at a set temperature.
4. Air quality station for Kathmandu valley. Problem: people discuss pollution without local data. Core: PMS5003 particle sensor, MQ-135, ESP32. Level: Medium. Viva value: calibration limits of cheap sensors. Twist: compare readings from two locations on one chart.
5. Cold-chain temperature logger. Problem: vaccines and milk must stay cold in transit. Core: DS18B20, SD card, RTC, buzzer. Level: Easy. Viva value: data logging, accuracy. Twist: SMS or push alert when limits are crossed.
Domain 2: Embedded systems
6. RFID access and attendance system. Problem: paper registers get forged. Core: RC522 RFID, Arduino or ESP32, RTC, SD card. Level: Easy. Viva value: SPI protocol. Twist: export a monthly CSV report automatically.
7. Fingerprint locker with tamper alarm. Problem: shared hostel lockers. Core: R307 fingerprint module, solenoid lock (12 V supply, separate from the MCU), accelerometer. Level: Medium. Viva value: authentication and false-accept/false-reject trade-off. Twist: three failed tries lock it for a time.
8. Smart energy meter (low-voltage demo). Problem: people cannot see which appliance wastes power. Core: INA219 and ESP32 for DC measurements. Level: Medium. Viva value: power (V × I) and energy (Wh). Twist: monthly cost estimate. Safety: demo with a low-voltage lamp; mains wiring is for a qualified electrician only.
9. Optical pulse and heart-rate display. Problem: students need a way to see how optical pulse sensing works. Core: MAX30102 pulse sensor, OLED. Level: Medium. Viva value: signal filtering. Twist: show a live pulse waveform on a learning prototype.
10. Custom data-acquisition board. Problem: lab instruments are expensive. Core: STM32 or ESP32, ADS1115 ADC, USB serial plotter. Level: Hard. Viva value: sampling, resolution, aliasing. Twist: compare your measurements with a lab multimeter and report error.
Domain 3: Robotics
11. Line follower with PID. Problem: simple on/off following is jerky. Core: Arduino, IR sensor array, L298N or TB6612 driver. Level: Medium. Viva value: PID tuning. Twist: plot error versus time for three tuning sets.
12. Disinfection or sprinkler rover. Problem: manual spraying of small farms. Core: Bluetooth or Wi-Fi control, pump, motor driver. Level: Easy. Viva value: motor drivers. Twist: add a tank-level cut-off.
13. Obstacle-avoiding wheelchair prototype. Problem: accidental collisions for users with limited control. Core: joystick, ultrasonic sensors, motor driver. Level: Medium. Viva value: safety design. Twist: a speed limit that falls as obstacles get closer. Use a scale model without a rider.
14. Robotic arm with gesture control. Problem: remote handling tasks. Core: servo motors, PCA9685, MPU6050 glove. Level: Medium. Viva value: kinematics basics. Twist: record and replay a movement sequence.
15. Autonomous mapping car. Problem: robots need to know where they are. Core: ESP32, wheel encoders, VL53L0X or lidar, serial output to a PC. Level: Hard. Viva value: odometry error. Twist: draw the room map on a laptop.
Domain 4: Power and renewable energy
16. Solar charge controller with MPPT. Problem: panels waste energy at fixed voltage. Core: buck converter, INA219, microcontroller. Level: Hard. Viva value: perturb-and-observe algorithm. Twist: compare against a direct connection.
17. Dual-axis solar tracker. Problem: fixed panels miss morning and evening sun. Core: four LDRs, two servos, Arduino. Level: Medium. Viva value: light-difference control. Twist: measure the actual extra energy collected, and report honestly if it is small.
18. Load-shedding auto-changeover indicator. Problem: manual switching between grid, inverter and solar. Core: voltage sensing with an isolated transformer, relays, display. Level: Medium. Viva value: isolation and switching delay. Twist: battery state-of-charge estimation. Safety: use low-voltage supplies for the demo.
19. Micro-hydro monitoring panel. Problem: village micro-hydro plants rarely have monitoring. Core: voltage, current and RPM sensing, ESP32. Level: Hard. Viva value: generator basics. Twist: scale it down with a small DC motor as generator.
Domain 5: Communication
20. LoRa village sensor network. Problem: no mobile coverage in some valleys. Core: two or more SX1276 modules, ESP32, gateway to the internet. Level: Hard. Viva value: range, spreading factor. Twist: measure range with line-of-sight and with obstacles.
21. Li-Fi text transmitter. Problem: show data over light. Core: LED, photodiode, Arduino. Level: Medium. Viva value: modulation. Twist: report bit error rate at different distances.
22. Earthquake alert relay. Problem: shaking is felt before people understand it. Core: MPU6050, ESP32, buzzer, MQTT. Level: Medium. Viva value: threshold detection. Twist: nodes share notifications after detecting local shaking. Be clear it is a prototype, not a certified system.
Domain 6: AI on the edge
23. Plant disease detector. Problem: farmers misread leaf spots. Core: ESP32-CAM or phone, small CNN (TensorFlow Lite). Level: Hard. Viva value: dataset, accuracy, overfitting. Twist: use leaves from local crops you photograph yourself.
24. Wake-word or voice-command switch. Problem: hands-free control. Core: ESP32 with I2S microphone, TinyML model. Level: Hard. Viva value: feature extraction. Twist: train on Nepali commands.
25. Vehicle counting for a junction. Problem: nobody measures traffic before building signals. Core: Raspberry Pi, camera, lightweight detector. Level: Hard. Viva value: precision and recall. Twist: compare counts with a manual count and show the error.
Scoring table: pick your project in five minutes
Score each idea from 1 to 5, then total. Pick the highest that you can actually finish.
Scroll sideways to see all columns.
| Factor | Ask yourself | Weight tip |
|---|---|---|
| Interest | Will I enjoy this at 11 pm in week 10? | Highest |
| Parts | Can I get the parts in Nepal within two weeks? | High |
| Theory | Can I explain every block on a whiteboard? | High |
| Demo | Does it work on a table in five minutes? | High |
| Novelty | Is there a twist the next group lacks? | Medium |
| Safety | Can I demo it without mains voltage? | Must be 5 |
If a project scores below 3 on Parts or Demo, simplify it now, not in month four.
Tips that save marks
- Build a minimum version first, then add the twist. A working basic project beats a half-finished advanced one.
- Keep a test log with real numbers: sensor error, response time, battery life. Examiners trust measurements.
- Prepare a block diagram, a circuit diagram and a flowchart before the viva.
- Know why you chose each component and one alternative.
- Report limits honestly. "This cheap sensor drifts in humidity" earns respect.
FAQ
Which project is best for a first-class result?
There is no single winner. Examiners tend to reward a clear problem, correct theory, measured results and a reliable demo. A well-documented medium project often beats an overambitious one.
Can two groups do the same idea?
Usually yes, but the twist, data and analysis should be your own. Ask your supervisor and always write your own report and code.
Do I need an expensive kit?
Most ideas above run on an ESP32 or Arduino-class board plus common sensors. Heavier ideas such as edge AI need a camera and sometimes a Raspberry Pi.
Are these ideas allowed in my university?
Requirements differ between departments and years. Check your department's format and have the idea approved before you buy parts.