Thesis & Design Report (Bachelor of Science in Electrical and Electronic Engineering)

Browse

Search Results

Now showing 1 - 10 of 575
  • Thumbnail Image
    Item
    A quadruped robot for real-time inspection and responsive operations
    (BRAC University, 2026-05) Hossain, Md. Shahariar; Abrar, Rafin Md; Shadid Al Akib; Ahmed, Tazwar; Rahman, Touhidur; Alam, Shahed; Oni, Atib Mohammad
    Industrial inspection in hazardous environments such as chemical plants, oil and gas facilities, power systems, and nuclear sites poses significant risks to human workers due to exposure to toxic gases, extreme temperatures, high voltage, and confined spaces. To address these challenges, this project presents the design and implementation of a quadrupedal robot capable of performing real-time inspection and responsive operations with minimal human intervention. The robot integrates a multi-sensor inspection framework for detecting structural and operational abnormalities, with particular focus on identifying surface cracks, microcracks or other early signs of equipment degradation. The system architecture combines a legged mobility platform, onboard computing unit, wireless communication module, and sensor suite including camera, ultrasonic, and IR-based sensing components. These subsystems enable the robot to collect and transmit real-time inspection data to operators for rapid monitoring and decision-making. In addition to routine inspection, the robot is also equipped with a manipulation mechanism for limited emergency operations such as operating switches or mechanical levers when direct human access is not feasible. The quadruped structure provides greater adaptability than conventional wheeled or tracked platforms in stairs, narrow passages, and uneven surfaces, making it more suitable for industrial inspection scenarios. This study addresses key limitations in existing inspection systems, including limited autonomy, infrastructure dependency, and high operational costs, by proposing a modular, cost-effective, and energy-efficient design. Experimental analysis demonstrates that the system can perform reliable inspections while reducing human exposure to hazardous environments. By enabling real-time crack and microcrack detection alongside general hazard inspection, the system supports predictive maintenance, reduces unplanned downtime, and improves worker safety. Overall, this work contributes a modular and practical robotic solution aligned with the growing industrial demand for intelligent, autonomous, and inspection-centric systems.
  • Thumbnail Image
    Item
    Smart whiteboard management system
    (BRAC University, 2025-10) Raiyan, Ekram Ahmed; Mercy Al Hossain; Mushfiq Us Salehin; Manjur, Md.; Rasheduzzaman, Mirza; Rahman, Md. Mosaddequr
    Our Smart Whiteboard Management System transforms the traditional classroom into a modern, interactive space. It's a portable, electric-powered system that cleans the board automatically using 99.5% of the surface area. A 5MP camera captures everything written--notes, equations, and diagrams--and converts them into editable LaTeX PDFs and shares them through a Flask-based web portal. The system also includes an AI chatbot tutor that provides instant help and reduces the need for human assistants by up to 50%. Using reed sensors and ESP32 microcontrollers, it selectively cleans only the required areas, making the process efficient and sustainable. With excellent digitization accuracy, it reduces paper waste and supports eco-friendly learning.
  • Thumbnail Image
    Item
    Early stage ML based non invasive breast cancer screening
    (BRAC University, 2026-01) Khan, Mohammad Fasiul Abedin; Nowshad, Farrdin; Nahean, Abrar Maksud; Mridul, MD. Abu Anas; Jahan, Nahid Akhter; Rasheduzzaman, Mirza; Rahman, Md. Mosaddequr
    Early breast cancer detection in low- and middle-income countries is limited by high screening costs, lack of infrastructure, and dependence on specialized facilities. This project presents a portable, low-cost, non-invasive AI-assisted breast cancer screening system using infrared thermography and machine learning, designed for deployment in resource-constrained settings. The system captures multi-view thermal images and analyzes temperature asymmetry and abnormal heat patterns using a convolutional neural network deployed on an embedded edge-computing platform. A structured engineering approach was followed, including evaluation of multiple design alternatives, optimization, sustainability, economic analysis, ethical compliance, and project management. The system provides an output, Benign or Malignant, to support clinical decision-making without replacing diagnostic procedures. The results demonstrate technical feasibility, affordability, and sustainability, establishing a strong foundation for IRB-guided clinical validation and scalable community-level screening.
  • Thumbnail Image
    Item
    Design and implementation of residential building management system
    (BRAC University, 2025-10) Khandaker, Md. Takky Hasan; Kabir, Mohammad Ibtehaz; Chowdhury, Fidaul Mowla; Siddique, Mohammad Samir; Hossain, A K M Shakhawat; Rahman, Md. Mosaddequr; Rasheduzzaman, Mirza; Rahman, Md. Mosaddequr
    This project presents the design and implementation of a Residential Building Management System (BMS) aimed at improving energy efficiency, safety and comfort in modern homes. The proposed system integrates IoT-based automation using the ESP8266 NodeMCU microcontroller and various sensors to monitor temperature, humidity, motion, fire and gas leakage. It allows real-time supervision and remote operation of lighting, HVAC, safety and security systems and can be controlled through the Blynk IoT platform. Centralized BMS was optimized for cost, performance and scalability in comparison to all three design approaches Distributed, Centralized and Hybrid. Our testing confirmed reliable communication, fast response and 15-25% reduction in energy consumption. The system minimizes power waste, ensures safety and keeps cost down that supports sustainability goals. In conclusion, the system shows an effective, scalable and ethical engineering solution for smart, sustainable residential environments in Bangladesh and similar urban areas.
  • Thumbnail Image
    Item
    Autonomous waste collecting device for Sugandha point of Cox's bazar sea beach
    (BRAC University, 2026) Uddin, Mohammad Riaz; Sreoshi, Sabrina Afrin; Hossain, Md. Shariar; Musabbi, Abdullah Al; Rahman, Dr. Mosaddequr; Islam, Mohaimenul; Bobby, Aldrin Nippon
    The issue of waste pollution of the coastal territories is also increasingly becoming a rising problem in Bangladesh, especially at the most popular sites like the Sugandha Point of the Cox's Bazar Sea Beach where tons of plastic and solid waste are becoming a daily occurrence and current means of clean-up is based mostly on manual labor. In this project, the proposal is to design and develop a cheap yet fully autonomous waste collector with the creation of a particular device designed to perform effectively in sandy and salty beaches. The proposed system will include an autonomous rover to identify typical wastes along with AI based computer vision, GPS based covering, ultrasonic and infrared sensors to avoid obstacles and a mechanical mechanism of waste collection in the form of a shovel. Once the waste is detected with the help of a camera and trained machine learning model, the rover is capable of autonomously picking a piece of garbage and placing it into an on-board container, which is emptied at a designated dumping point once full. Various design solutions have been considered, such as a robotic arm and a conveyor belt system, and the shovel mechanism has been chosen because of the simplicity of the mechanism, its lasting capacity, energy- efficiency, and its application in non-even sandy areas. There are experimental findings to prove that the device can be effectively used to clean a predetermined beach area with reducing the cost of operation and less human intervention. The above system can serve as a viable solution to sustainable beach cleaning, and it can place autonomous robotics in an environmentally-sustainable context, especially in developing nations.
  • Thumbnail Image
    Item
    IOT based centralized net meter monitoring and management system for grid connected solar PV
    (BRAC University, 2026) Mujahid, Md. Mustaq; Zuboraz, Md. Zubayet Hossain; Iqtiad, Ajmain; Hasan, Md. Shakil; Khan, Dr. Shahidul Islam; Rasheduzzaman, Dr. Mirza; Rahman, Md. Mosaddequr
    The integration of rooftop solar photovoltaic (PV) systems through net metering has significant potential to improve energy sustainability in Bangladesh; however, the lack of intelligent monitoring and control often leads to inefficient utilization of installed systems. This project presents the design and hardware implementation of an IoT-based centralized net meter monitoring and management system for a grid-connected solar PV battery setup. The ESP32 microcontroller is a component of the system that measures real-time electrical parameters (voltage, current, power, energy import/export, and battery state-of-charge). Control logic based on relays is used to have automatic control over grid import, grid export and battery operation. The visualization of system data is provided with an IoT dashboard to give a clear understanding of net metering behavior. A hardware prototype was created and tested in order to confirm the system performance. The feasibility was also assessed by carrying out an economic analysis based on cost breakdown, payback period and sensitivity analysis. The outcomes showcase a better visibility of energy, efficiency in operations and realistic use as far as rooftop solar systems are concerned.
  • Thumbnail Image
    Item
    Design and implementation of a weather station for flood-prone agricultural lands
    (BRAC University, 2026-01) Mohiuddin, Akif; Mojumder, Arfanul Hoq; Inshaniate, Brinto; Sakib, Tanjil Abrar; Rahman, Md. Mosadeqqur; Islam, Mohaimenul; Bobby, Aldrin Nippon
    One of the frequent natural hazards in Bangladesh is flooding which has a devastating impact on the agricultural productivity especially in the low-lying and rural areas. The absence of localized, real-time weather surveillance and the early notification of floods frequently causes a lot of crop loss, economic setbacks, and a decrease in livelihood of the agricultural people. This project will constitute the design and deployment of a low-cost, IoT based weather station that will be specifically designed to work on flood prone agricultural land. The system, proposed, incorporates the various environmental sensors to measure the significant parameters of rainfall, temperature, humidity, soil moisture, wind speed, atmospheric pressure, and water level in real-time. To process, store, and analyze data, data is sent via long-range and low-power LoRa communication network to a cloud-based platform. The prediction model application is based on machine learning to analyze real-time and historical data and estimate flood risks to provide early warnings. The system runs off a solar-powered energy management unit which has battery backup thus making it operate continuously in off-grid and remote regions. The system sends real-time alerts via SMS and a user-friendly web dashboard to the farmers. Experimental analysis proves to have credible data transmission, reasonable sensor accuracy, power-saving, and useful flood risk forecasting. The proposed solution provides a cost-effective, energy-efficient, and scalable solution to enhance flood preparedness, loss of crops, and sustainable agricultural practices in flood prone areas of Bangladesh.
  • Thumbnail Image
    Item
    TorqueSAM: unsupervised kidney CT analysis with localization and SAM-integrated torque clustering segmentation
    (BRAC University, 2026-02) Rahat, Shahzalal Khan; Apurbo, Shah Miran; Pranty, Anika Tazin; Yesmin, Mahejabin; Biplob, Md. Asif Hasan; Alam, Md. Golam Rabiul
    Chronic kidney disease is a growing public health hazard that often leads to permanent renal system failure if not diagnosed and treated by experts within a treatable time interval. In addition to that, the shortage of expert nephrologists in unprivileged areas slows down the early detection and prevention. However, AI driven solutions can play a vital role in ensuring early detection and prevention despite the shortage of experts in many cases. But most existing AI driven automated renal disease detection systems heavily rely on manually annotated data, which still needs the experts’ intervention to detect any chronic renal illness. To address this issue, this study proposes and tests a fully unsupervised method to detect one critical chronic renal disease: Renal stone, often interpreted as a stone that does not rely on any annotated data. In this regard, our proposed method, namely TorqueSAM, introduces a different approach to work with modern detection and segmentation models, along with a classification module to detect any Renal Stone without relying on any ground truth labels. Our approach is tested on a dataset of 6,454 axial and coronal CT scan images. Our study focuses on one alarming illness that is highly dominant and associated with long-term kidney failure. TorqueSAM shows that the unsupervised image segmentation and classification surpass the classical supervised approaches significantly in terms of reducing the annotation overhead, time, and resources. For renal stone, TorqueSAM showed dice scores of 0.5767, IoU of 0.4183, precision of 0.9296, recall of 0.4655, accuracy of 0.9992, and specificity of 0.9998. However, the classification results of the clustering is found to be 0.8232 and 0.8472 for ARI and NMI, respectively. Despite having some limitations in the segmentation, this research shows the potential to outperform the classical supervised techniques, prioritizing the optimization of the region of interest (ROI) to enhance the predictive accuracy while reducing annotation overhead, which allows TorqueSAM to be a scalable solution to mitigate the sufferings of renal stone detection.
  • Thumbnail Image
    Item
    Design and implementation of an indoor hydroponic system for enhanced crop production
    (BRAC University, 2026) Tuli, Mahabuba Akter; Hasan, Md. Samir; Islam, Md. Sazzadul; Shakib, Sadman; Bhuian, Mohammed Belal Hossain; Islam, Md. Mahmudul; Ahmed, Sabbir
    Urban agriculture in densely populated regions, which faces severe constraints due to limited arable land, soil degradation, and water scarcity, requires efficient and space-saving cultivation solutions. This project introduces an automated Indoor Vertical Hydroponic System specifically designed for space-constrained urban environments. By incorporating IoT sensors that monitor TDS, turbidity, water level, temperature, and humidity, the system allows for precise environmental tracking to automate nutrient delivery and optimize crop health. The system operates largely independently of outdoor weather by utilizing spectrum-controlled LED lighting and a camera module for real-time visual monitoring, ensuring year-round production. The use of automated water and oxygen pumps guarantees efficient resource circulation, significantly minimizing water waste compared to traditional soil-based farming. Moreover, the system is designed as a scalable, ready-made agricultural product, providing a user-friendly solution that maximizes yield per square foot. By addressing both space constraints and resource efficiency, this project not only improves the competitiveness of the agricultural sector but also promotes sustainable food security in Bangladesh's land-scarce markets.
  • Thumbnail Image
    Item
    Psychmatrix: empowering users through advanced interactive control
    (BRAC University, 2025-10) Rifat Bin Reza; Ahmed, Shadman; Afrin, Barida; Mrida, Imtiaj Saiem; Khan, Shahidul Islam; Zunaed, Mohammad; Chowdhury, Saad Mahbub
    This project introduces the concept of PsychMatrix, a very simple yet effective and affordable assistive technology for controlling home appliances using hand gestures without touching the appliance. The system makes use of wearable gloves embedded with flex sensors to obtain information related to the bending of the fingers and muscles. These signals are processed by ESP32 microcontroller which sends commands to a cloud-based firebase realtime database via Wifi . These commands are then sent to a second ESP32 unit which works as a home controller by turning appliances on and off using relays. An assistant mobile application gives the system real-time monitoring and user interface. PsychMatrix is a control board with a safe, responsive design (with a measured end-to-end Latency of under 500ms) arranged to benefit the elderly and those with disabilities, as well as high-voltage workers, which has an interaction recognition accuracy rate of 98%. The system proves that with low-cost off-the-shelf components, a well-built, cloud-connected assistive technology can be enacted to ensure a higher degree of independence and safety.