To address Bangladesh’s ongoing load shedding and power planning issues, we built a deep learning–based forecasting system using five years of BPDB data, enriched with temperature, holiday, and seasonal features. After training LSTM, GRU, BiLSTM, and BiGRU models, the BiGRU architecture achieved the highest accuracy (4.75% MAPE) and was used for both national and division-wise predictions. The system enables proactive demand forecasting and has the potential to reduce planning inefficiencies and energy waste.

Teammates: Tasnia Nafs, Sakif Yeaser
Project Report

Built a smart oven prototype using the AT89C51 microcontroller, featuring keypad-based timer input, real-time countdown via 7-segment display, and LCD-based status messages. Included safety features like an emergency stop (RESET) and buzzer alerts. LEDs indicate oven status (running, finished, powered), making it a compact, user-friendly embedded system for controlled heating tasks. 

We developed an intelligent temperature-adaptive motor control system using the AT89C51 microcontroller in conjunction with the ADC0804 analog-to-digital converter and the L293D motor driver IC. This project dynamically adjusts motor (fan) speed based on real-time temperature data, mimicking PWM behavior through software-driven ON-OFF cycles. The system allows users to configure multiple temperature thresholds—minimum, medium, maximum, and critical—via a keypad interface. During operation, the LCD screen continuously displays the current temperature and fan status, while visual indicators (LEDs) represent the selected speed level. If the temperature surpasses the critical limit, the system automatically halts the motor and activates a buzzer alarm to prevent overheating. Designed for simplicity, affordability, and efficiency, this microcontroller-based solution demonstrates real-time environmental responsiveness suitable for PCs, small enclosures, and temperature-sensitive electronic systems. 

Teammates: Sadat A. Rashad, Zerin Yeasmin, Raiyan Ibne Hossain, Sarjo Sonko
Project Report

Our team developed an interactive MATLAB App titled "Signals and Systems Playground" as part of our Signals and Systems Lab coursework. The app is designed to simplify key concepts of signals and systems, offering modules for signal visualization, image processing, and audio processing.  

The signals module allows users to visualize various signal types such as sinusoidal and exponential signals and perform operations like shifting, scaling, and time-scaling. In the image processing module, users can resize, crop, and adjust the contrast of images, providing hands-on experience with image manipulation techniques. The audio processing section enables users to analyze waveforms, adjust gain, and remove silence in audio files, demonstrating practical audio signal handling.  

This project gave us the opportunity to enhance our MATLAB programming skills and apply theoretical knowledge to develop a functional and versatile learning tool for both students and educators.  

Teammates: Tasnia Nafs, Muslima Siddiqui
Project Report
Download app

Our team worked on an innovative project to develop a Real-time sign language recognition system as part of our Electrical and Electronic Workshop coursework. The system aimed to enhance communication accessibility for the Deaf and hard-of-hearing communities by translating sign language into text and speech.

The project involved training a Transformer-based model on curated datasets such as the American Sign Language Dataset and WLASL. The model was deployed using OpenCV and  TensorFlow Lite for real-time video detection, enabling the system to recognize gestures accurately. Additionally, the system featured text-to-speech functionality for seamless communication.

One of the significant challenges was handling data variations due to webcam inconsistencies, which we mitigated through robust preprocessing techniques. We also relied on pre-existing datasets due to the lack of expertise in collecting sign language data.

This project allowed us to explore advanced machine learning concepts, improve our programming skills, and develop a deeper understanding of accessibility solutions. The result was a functional system that bridges the communication gap for sign language users and non-signers.

Teammates:  Tasnia Nafs , Omar Faruk, Sanzida Tasnim Nisa
Project Report

As part of our Power Electronics coursework, our team developed an adjustable timer-based smartphone charging circuit aimed at enhancing battery lifespan and improving energy efficiency. The project focused on creating a user-friendly solution to optimize charging processes while safeguarding mobile phone batteries from overcharging.

The circuit design utilized a 555 timer IC and a relay to enable precise control over the charging duration. Users could easily set adjustable timers ranging from 8 minutes to 1.5 hours, tailoring the charging time according to their specific needs. This ensured that the battery received just the right amount of charge, preventing overcharging and reducing wear on the battery over time.

This project allowed us to explore practical circuit implementation while addressing real-world problems in mobile device usage. The final product was a reliable and efficient charging solution that contributes to extending battery life and promoting energy-saving practices.

Teammates:  Tasnia Nafs , Omar Faruk, Sanzida Tasnim Nisa
Project Report

Our team, "Inception," embarked on an exciting project as part of our Digital Logic Design course. Our goal was to design and implement a simple 8-bit architecture-based computer using Proteus software. The computer consisted of various critical components, including an Input unit, Output unit, RAM, Memory Address Register, Arithmetic Logic Unit, 8-bit bus, Address Selector, Counter, and Clock Pulses Generation (both manual and auto-generated).

One of the key challenges we faced was to reduce the T-state from 6 to 4, which we successfully achieved through our dedication and hard work. We then brought the circuit to life by implementing it on a breadboard using standard ICs. To further enhance our project, we created Printed Circuit Boards (PCBs) for each module and assembled the entire project.

This project allowed us to gain valuable hands-on experience in both software and hardware design, as well as a deeper understanding of theoretical concepts. We learned about the design process, the importance of collaboration, and the practical applications of digital logic. The end result was a fully functional 8-bit architecture-based computer that demonstrated our technical skills and teamwork abilities.

Overall, this project was a fantastic learning opportunity that will benefit us in our future careers. We are proud of what we achieved and are eager to tackle new challenges in the future. 

Teammates : Tasnia Nafs, Shahriar Sadi, Sanjida Tasnim Nisa, Md. Mohibul Haque
Project Report

Developed a simple GUI in MATLAB as part of the Open-Ended Lab Project for the MATLAB course, creating an interactive platform to streamline complaint management for the IUT Hall of Residence. The application enables students to log issues related to services like carpentry, plumbing, electrical, internet, and cafeteria, featuring touchscreen accessibility, automated data logging, dynamic forms, cloud integration, and enhanced usability. 

Teammates : Tasnia Nafs, Muslima Siddiqui
Project Report
Download app

During my first year as an undergraduate student, my team and I built several soccerbots and a line follower robot. We participated in various competitions nationwide, gaining valuable hands-on experience in robotics, automation, and electronics. These projects laid the foundation for our skills in these fields and provided us with an early edge in pursuing advanced academic endeavors, such as joining the Mars Rover Team and contributing to other significant projects.

Teammates : Tasnia Nafs, Junayed Ahmed