Projects
Mine Sweeping Robot (Raspberry Pi)
Developed a landmine-detecting robot using a Raspberry Pi and sensors as part of a national Robo-Fight competition at Malaysia. The system uses obstacle avoidance and metal detection to simulate real-world mine detection and removal.
You can see the robot demo and look at my role 2:03.
GitHub repository (with teammate Md. Kaiser Raihan): 🔗 View on GitHub
Collaborated with: Rahi008
Fighting Robot for Robo-fight Competition (Python)
Developed a robot with competitive fighting capabilities controlled via Python programming.
Car Speed Measurement Device (Microcontroller)
Year: 2017 | Institution: Chittagong University of Engineering & Technology (CUET)
This was my 3rd-year academic project where I designed and implemented a car speed measurement system using a microcontroller-based embedded setup.
The device uses two infrared (IR) sensors placed at a fixed distance (e.g., 20 cm apart) to detect the movement of a toy or model car. As the car passes each sensor, the microcontroller captures timestamps using its internal timer functions. The speed is then calculated with the formula Speed = Distance / Time, and the result is shown in real-time on a 16x2 LCD display.
Beyond detecting speed, the system introduced me to real-time processing, sensor interfacing, digital electronics, and embedded programming. It required precise timing mechanisms using interrupts and the integration of both hardware and software components.
Key Features:
- Real-time speed calculation using two IR sensors
- Microcontroller-based timing and interrupt handling
- Display of speed in km/h on LCD
- Compact and low-cost hardware design
- Built using Arduino Uno, Embedded C, and standard IR modules
Arduino-based Automatic Plant Irrigation System
Year: 2015 | Institution: Chittagong University of Engineering & Technology (CUET)
Designed and developed an automatic plant irrigation system using an Arduino Uno and a resistive soil moisture sensor. The project utilizes a voltage divider circuit to convert varying soil resistance into analog voltage, which the Arduino reads to determine moisture levels.
When the soil is too dry (i.e., the analog voltage falls below a set threshold), the Arduino activates a water pump via a relay module. Once the soil reaches adequate moisture, the pump is automatically turned off — ensuring optimal water usage and preventing overwatering.
This was one of my first hardware automation projects and helped me understand analog sensor behavior, voltage divider principles, and relay-based actuation. It was a significant milestone in developing practical embedded systems knowledge.
Key Features:
- Soil moisture monitoring using voltage divider principle
- Analog input processing on Arduino Uno
- Relay-controlled submersible water pump
- Power-efficient, self-regulating system
EMG Signal Analysis and Healthy Person Detection (MATLAB)
Analyzed EMG signals to differentiate between healthy and abnormal muscle activity.
Implementation of Machine Learning Classifiers (Python)
Implemented KNN, Logistic Regression, Kernel SVM, and GMM for data classification.
Optimization Schemes for Stochastic Gradient Descent (Python, PyCharm)
Explored optimization techniques for improving SGD performance in deep learning models.
Step-by-Step Implementation of Convolutional Neural Network (Python, PyCharm)
Developed a CNN model from scratch to understand convolution, pooling, and activation functions.
Banking Distress Prediction via Machine Learning (MATLAB)
Applied machine learning techniques to predict financial distress in banks.
Design and Analysis of n-bit Comparator in Cadence Virtuoso (VHDL)
Designed and simulated an n-bit comparator circuit using Cadence Virtuoso, covering schematic, layout, DRC/LVS checks, and Verilog simulation. Focused on propagation delay, power, and area optimization.
Year: 2018
Tools used: Cadence Virtuoso, Verilog, SimVision