CSE Portal — Boot Sequence
>C5@$_# ]>--<90B AC2 >&D\$7
······· ········ ·······
········· ·········· ·······
········· ······· ·····
]&}#+0]\ B%<8^B$ 92 !*/8
······· ·········· ··· ····
········ ····· ·· ········
········· ······· ·······
Project ID = S25SDP 01
Supervisor: Khalid Abualsaud
Asmaa Alansari, Bashaer Alkuwari, Hissa Al-Ali, Jawaher Al-Mohamed
Stress is a significant public health concern in Qatar and the Gulf region. Autism Spectrum Disorder (ASD) affects 1.14% [1] of school-aged children, which is estimated to be 1 in 87, which is higher than the global average rate of 1 out of 100 [14] in Qatar. Research shows that (DMS-5), people with ASD have difficulties in communication and social interaction because of atypical information processing and sensory integration abnormalities. Stress can impact the physical and mental health of a person with ASD. People with ASD may be at high risk of experiencing very stressful and traumatic events, which can negatively affect mental health. According to DSM-5 [10], approximately 70% of people with ASD have a comorbid mental health disorder, and up to 40% may have two or more disorders [11]. Usually, people with ASD have problems related to sensory processing; thus, they could experience a sensory overload, in which one or more senses react to stimuli, which can trigger elevated stress levels. Existing wearable devices have shown potential in monitoring physiological signs such as heart rate, skin conductance, and body temperature, which are commonly associated with stress. However, current technologies do not fully address the broader context of stress triggers for individuals with autism. Specifically, these devices lack sensors and algorithms capable of detecting environmental triggers like social interactions, light, and sound, which shows gaps in detecting stress in individuals with ASD. This project aims to bridge these gaps by developing a Smart Wearable Stress Management System tailored to the needs of children with autism in Qatar. The proposed device will integrate lightweight algorithms designed to track stress levels in real-time and identify threshold peaks associated with unique stress patterns, such as sensory overload and stimming behaviors. Additionally, the system will offer adaptive responses and provide immediate stress relief through calming stimuli like soothing sounds or light therapy. The wearable device incorporates sensors capable of detecting environmental stressors, such as social interactions, light intensity, and noise levels, which influence stress levels. Also, it offers personalization options like color choices and nature sounds (e.g., birds, water, rain) to accommodate sensory preferences. The device user interface will be sensory-friendly, using materials that are comfortable for extended wear, and allow easy customization of alerts and responses based on the user's sensitivity. The system will also enable caregiver connectivity, providing real-time notifications to facilitate immediate intervention.
Project ID = S25SDP 02
Supervisor: Hela Chamkhia/Muhammed Azeem (Co-supervisor)
Ghalya Al-Jobara, Lulwa Almalki, Lolwa Taymour, Wejdan Al-Marri
The Smart School Bus Tracking & Attendance System addresses critical safety concerns in student transportation by implementing an integrated technological solution. This project develops a comprehensive system that monitors students as they board and exit school buses, tracks bus locations in real-time, and provides immediate notifications to parents and school administrators. By utilizing facial recognition, RFID technology, GPS tracking, and a mobile application, the system creates a secure transportation environment where all stakeholders can access relevant information based on their roles. In addition to monitoring student seating, detecting unauthorized individuals, and automatically recording attendance, the technology also promptly notifies students of impending bus stops and makes sure no kid is inadvertently left behind. By integrating technology, schools and parents may communicate more effectively, streamline attendance processes, increase student safety, and ensure accountability throughout the transportation process. In order to implement the project, physical elements like cameras and RFID readers are combined with a strong software architecture that includes backend processing, database administration, and an intuitive mobile interface. The end result is a dependable system that gives parents and school officials ease and peace of mind while addressing the safety issues of student transportation.
Project ID = S25SDP 03
Supervisor: Muhammad Arsalan
Almaha Al-Kaabi, Amna Al-Obaidli, Duha Bukshaisha, Noor Al-Mahmoud
Increasing demands for assistive technologies have favored rapid development in the design of intelligent and autonomous systems aimed at improving users' safety, mobility, and independence. Because of this need, this paper presents a Smart Assistive System for the Visually Impaired developed using Vuzix Blade 2 smart glasses, a Raspberry Pi 4, and an Android-based mobile application. The system is designed to allow visually impaired people to increase their awareness of the environment through its three main core features: face recognition, medicine identification, and SOS alerts in emergency situations. All recognition features are executed completely offline without cloud services, while the only function that requires internet connectivity is SOS. It uses a face recognition module featuring a MobileNetV2 deep-learning model trained in PyTorch to recognize people in real time. With a single tap on the touch surface of the glasses, the camera takes a photo, sends the image instantly to the Raspberry Pi over a private Wi-Fi network, and sends the inference results back instantly. If it detects an unknown person, the system triggers an email alert with the captured image to predefined emergency contacts. The result of the recognition is also read out loud with TTS audio feedback, like "Known person" or "Unknown person." A double tap invokes the medicine recognition feature, which uses OCR via OpenCV and Tesseract to extract and interpret printed text from medicine packages. It will then take this extracted text and match it against a controlled, whitelisted set of recognized medicines, such as Panadol, Brufen, and Claritin, using fuzzy string matching to account for imperfect lighting or rotations. The name is then audibly announced to the user for safe medication handling. The triple tap SOS feature triggers the NEO-6M GPS module connected to the Raspberry Pi via UART. This captures an image, retrieves the user’s GPS coordinates, and sends an emergency email with a Google Maps link and the photo attached to the contacts. This allows immediate alerts in emergency situations, where location and visual context of the user can be shared securely. The system operates fully offline through the self-hosted Wi-Fi access point created by the Raspberry Pi, which provides always reliable communication between the glasses and the server. All captured data, predictions, and overlays are stored locally on the device, ensuring privacy and independence of internet connectivity. In sum, this project presents an integrated assistive solution in one wearable device, developing the intersection of artificial intelligence, computer vision, and embedded systems. The novelty is the use of gesture-based interaction, real-time offline AI inference, and GPS-based emergency alerting to create an inclusive, privacy-preserving platform for visually impaired users. This, in turn, yields a scalable and efficient system that improves safety, autonomy, and confidence in daily navigation.
Project ID = S25SDP 04
Supervisor: Uvais Qidwai
Duaa Ahmad, Khadija Ashfaq, Yusra Bashir
Petriscope is a smart, portable microscope system designed to make the process of analyzing zebrafish eggs faster, easier, and more accurate. Traditional methods of checking the condition of eggs are often slow and require manual work, which can lead to mistakes. Our project aims to solve this by using a Raspberry Pi 5 along with a camera and a touch screen to automatically take images of the eggs and classify them as alive, dead, or dying. The system uses image processing techniques developed in MATLAB, and the main control code is written in Python using the Thonny IDE. The touch screen helps users interact with the system and view the results easily. Petriscope is designed to be low-cost, easy to use, and suitable for both labs and fieldwork. By reducing manual effort and increasing reliability, it provides a useful tool for researchers, students, and professionals. The combination of hardware and software in this project shows how technology can help improve accuracy and efficiency in scientific research.
Project ID = S25SDP 05
Supervisor: Mohamed Al-Meer
Amna Al-mohannadi, Noof Al-Hail, Reem Alkorbi, Noora Al-Ibrahim
Smart Vision Testing System addresses the limitations of traditional manual vision-testing methods, such as the Snellen chart, which depend on human observation and require a medical specialist. These limitations create challenges for patients with communication difficulties and for communities with limited clinical access. This project presents a digital and accessible alternative using a hybrid hardware–software architecture. The system employs an ESP32 microcontroller for patient input and a Raspberry Pi 4/5 for visual output and processing through a Python-based graphical interface. Communication between devices is handled wirelessly using the WebSocket protocol. Two testing modes are supported: keypad-based Snellen test and a joystick-controlled E-chart test for users who cannot read letters. The Raspberry Pi evaluates responses in real time and automatically generates a performance report at the end of each session. By automating evaluation and reducing human error, the system improves accuracy, consistency, and usability. The prototype demonstrates an affordable and inclusive approach suitable for clinics, schools, and remote environments.
Project ID = S25SDP 06
Supervisor: Muhammad Arsalan
Danah Zeid, Hend Abdulhameed, Masah Alrefaii Aboulabada
In high-security applications, single-factor authentication methods such as keys or RFID cards are increasingly inadequate for ensuring secure access to restricted areas. This paper presents the design and implementation of a multi-layered, biometric-based access control system that enhances security, privacy, and operational efficiency. The proposed system utilizes three sequential authentication layers: fingerprint recognition, facial recognition, and passcode validation, all executed on an edge artificial intelligence (AI) platform (NVIDIA Jetson Nano) for real-time, offline processing. An Arduino Uno microcontroller interfaces with a fingerprint sensor and controls access to a secured corridor via a micro servo motor, transmitting authentication status to the Jetson Nano. The authentication sequence initiates with fingerprint verification at the entry point. Upon success, region-of-interest (ROI)-based facial recognition is performed in a monitoring corridor. A final numeric passcode validation stage acts as a fallback mechanism in the event of biometric failure. Successful authentication activates a role-based access control (RBAC) interface that dynamically displays authorized access points. The Jetson Nano then controls solenoid-actuated door locks via relays based on the user’s role. Repeated authentication failures trigger an alert mechanism to notify security personnel. By combining biometric and knowledge-based authentication methods, the system provides high accuracy, redundancy, and spoofing resistance. This work demonstrates a cost-effective, scalable, and user-centered approach to secure access control, contributing to the advancement of embedded systems in physical security applications.
Project ID = S25SDP 07
Supervisor: Mohammed Al-Sada
Aisha Al-Naimi, Reema Al Bouainain, Sundus Al-Qadi
While museums serve as intriguing venues for learning, the majority of tours are perceived as monotonous, particularly by younger audiences. A significant number of museums employ human guides, which can be both costly and occasionally unavailable. Given the restricted opportunities for interaction or inquiry, tours may come across as passive and unilateral. To enhance the appeal, inclusivity, and accessibility of museums for all individuals, it is imperative to address this concern. Our project proposes the development of a mobile museum tour robot as an effective solution to this particular issue. Through engaging in conversation, responding to guests, and autonomously navigating, this robot can enhance interactivity. It incorporates a holographic display that is designed to vividly bring history to life in a visually captivating manner. Unlike human guides, the robot is perpetually accessible and capable of conducting tours continuously without the requirement for shifts or breaks. This innovation may assist museums in Qatar in attracting a larger number of visitors, conserving time and effort, and diminishing their reliance on human staff by providing a sophisticated, high-technology experience. Furthermore, the robot is intentionally designed to respect Qatari customs and values and to demonstrate cultural awareness. Despite the achievement of several significant milestones, including the development of the robot’s AI speech processing and conversational capabilities, certain technical challenges remain to be addressed. These challenges encompass real-time data processing, the integration of holographic displays, and effective navigation and obstacle avoidance. Moreover, we are in the ongoing process of collecting precise museum data, formulating an interaction style for the robot that fosters visitor trust, and ensuring that the robot adheres to Qatari social and cultural norms. Not withstanding these persistent tasks, this project provides an insight into the future of interactive learning and tourism, representing a substantial advancement toward the establishment of a smart, culturally attuned museum tour experience in Qatar.
Project ID = S25SDP 08
Supervisor: Junaid Qadir
Aisha Al-Mannai, Fatima Saleh, Naema Al-Abdulla, Noora Al-Jasmi
This project addresses the challenge of developing critical thinking and problem-solving skills, which are essential across various sectors, especially for decision-makers. These skills play a critical role in education, design projects, and research. However, they are often acquired through experience rather than structured learning methods, making their development inconsistent. The lack of these skills can lead to poor decisions that negatively affect outcomes. To address this challenge, we developed an advanced GPT-4-based brainstorming assistant that applies Edward de Bono’s Six Thinking Hats framework through a multi-agent system architecture. Each “hat” is controlled by a dedicated agent with a specialized system prompt, coordinated by a supervisor agent to ensure logical flow and coherence. This design enables more targeted and organized critical thinking support for students during their early design phases. The system is implemented as a full-stack web application. The front-end is built with React.js, while the backend is deployed on a serverless platform using Next.js and Supabase. We integrated a Retrieval-Augmented Generation (RAG) feature, allowing the chatbot to semantically process uploaded PDF project files and incorporate relevant content into its responses. This enriches the quality of advice provided and supports more personalized user guidance. Furthermore, the system applies advanced prompt engineering strategies such as role-based prompting, chain-of-thought reasoning, and question-based clarification iteratively refined through student feedback and testing. We evaluated both non-RAG and RAG-enabled versions of the chatbot through student surveys and performance metrics to assess clarity, structure, and brainstorming support. The uniqueness of our design lies in its ability to deliver structured, less biased feedback to users, supporting comprehensive and reflective brainstorming. Ultimately, this tool is positioned as a valuable assistant for students working on Senior Design Project (SDP) that requires critical thinking and problem-solving tasks. The system aims to reduce cognitive bias in decision-making by enforcing balanced perspectives through the Six Thinking Hats method. Each hat focuses on a distinct mode of reasoning, ensuring that no single viewpoint dominates the conversation. This structured approach leads to more objective and comprehensive brainstorming.
Project ID = S25SDP 09
Supervisor: Mohamed Al-Meer
Aisha Al-Assiri, Amna Saifaldeen, Maryam Al-Jabir, Roudha Alkhalaf
The objective of the Face Recognition Door Lock project is to improve the safety of homes and small businesses by utilizing facial verification that is not prohibitively expensive. For the purpose of doing real-time local facial recognition without relying on cloud services, the system makes use of a Raspberry Pi for its implementation. As a result, the user's privacy has been protected. Among the most important features are the automated unlocking of doors for known users and the refusal of entry to faces that have not been registered. Through the use of real-time testing, the project successfully established its ability to provide reliable functioning, with a high level of accuracy in recognizing authorized users. A market survey indicated that there is considerable demand, particularly among young individuals who are proficient in technology and who are experiencing problems such as key loss. The system is distinguished from expensive market goods that commonly rely on cloud storage owing to the fact that it is inexpensive (estimated to be between $100 and $200) and easy to set up. After being successfully evaluated in a number of different user scenarios, the solution was able to demonstrate its dependability, security, and potential for future integration into smart home systems.
Project ID = S25SDP 10
Supervisor: Abdelkarim Erradi
Aisha Ali, Samia Hasan, Zunaira Hamid, Zobia Zia
"While the Quran holds global significance, many of its readers rely on translations and Tafsir for comprehension. A key limitation in current digital Quran tools is the common lack of topic-driven Tafsir features, restricting users' ability to synthesize related verses and grasp broader thematic concepts effectively. Understanding the Quran's thematic structure within individual Surahs remains challenging with current digital tools. This project introduces a Quran application designed to bridge this gap via an interactive 'Topics Map'. This feature links verses addressing the same subject within a single Surah, utilizing digitized versions of At-Tafsīr al-Wajīz الوجيز and At-Tafsīr al-Wasīṭ الوسيط (from https://quranok.com/ lead by Prof. Abdulsalam Almajeedy). The goal is to provide a searchable and educational tool that illuminates a Surah's internal coherence, thematic development, and verse arrangement, thereby facilitating a more contextually grounded exploration of the Quranic text. Key contributions of this project include:
Project ID = S25SDP 11
Supervisor: Wadha Labda
Mathayel Al-Hawal, Najlaa Al-Sahel, Raghad Saleh, Sham Alkhais
University students and professors have to deal with multiple courses, keeping track of assessments, and organizing workloads every semester. Students often miss deadlines, struggle with time management and experience academic stress. The instructors also spend considerable time reminding students about deadlines, syllabus updates, and keeping assessment dates straight. Existing LMSs, such as Blackboard, provide course materials, but lack automation of personalized organization, progress monitoring, and workload visualization. This lack indicates a deep need for an intelligent approach that will help learners and educators improve academic planning, monitoring, and communication. TeachMate is an AI-enhanced web-based platform designed to address these challenges by automating assessment extraction, workload visualization, and academic progress tracking. The system utilizes Natural Language Processing (NLP) to extract assessment details directly from uploaded syllabi and includes modules for automated deadline management, weekly study planning, notification scheduling, and cross-course workload balancing. Students get a personalized dashboard, monthly task checklist, discussion threads, and AI-generated weekly study plans. Additionally, instructors are provided with assessment editing tools, modification history tracking, cross-section workload heatmaps, conflict advisories, and advanced analytics dashboards to help drive decisions. TeachMate uses NLP and AI techniques for automatically extracting assessment data and resources from uploaded syllabi, scheduling reminders according to preparation time, and generating performance analyses and workload recommendations. Achievements include developing a responsive and accessible interface with real-time notifications and discussion threads, integrating AI modules for automated planning, alerts, and analytics, and designing a scalable, optimized database to support real-time assessment updates and analytics generation. Other features include cross-section analytics, detection of conflict algorithms, creation of reports in PDF and CSV format, and comprehensive testing for all functional modules. Overall, TeachMate is a singular contribution to academic technology, bringing AI-assisted planning and workload analytics together for the first time in one system.
Project ID = S25SDP 12
Supervisor: Saeed Salem
Menatalla Abdulhamid, Maryam Alsheikh, Razan Alchikh, Taleela Al-Muhannadi
Community initiatives play a vital role in promoting social responsibility, managing societal demands and encouraging teamwork. However, coordinating initiatives and managing these opportunities around the world continues to be challenging for both individuals and organizers due to the lack of centralized volunteer management platforms. Consequently, individuals suffer in managing, coordinating and monitoring activities. Addressing this issue is crucial in encouraging a culture of community driven initiatives and enhancing their overall impact. Our project, Bussma, is a mobile application that enables organizers to initiate initiatives and monitor the progress of executing humanitarian and society-driven campaigns. The tool is primarily geared towards non-governmental organizations that execute humanitarian campaigns and food drives. The platform allows initiative managers to create opportunities, enlist volunteers and employees to tasks, provide real-time analysis on the progress of each initiative using an interactive geospatial visualization. The map dynamically updates, enabling users to monitor progress and recognize accomplishments. Additionally, the application includes features such as tracking task progress, visualizing contributions and proximity notifications to encourage engagement. Bussma seamlessly monitors real-time monitoring of projects and provides a user-friendly interface that makes initiative managing and creation easier, and a rewarding experience that encourages ongoing participation are some of our main achievements. Our application increases social impact, improves initiative outreach, and establishes a clear and accessible system for tracking contributions by simplifying initiative management. The features provided in Bussma are unique and the platform enables the integration of all steps in one app. In conclusion, this project not only addresses the need for better task coordination but also strengthens community-driven initiatives, ultimately contributing to a more socially responsible society across the world.
Project ID = S25SDP 13
Supervisor: Tamer Elsayed
Nedaa Al-Jabri, Shama Al-Ahmed
In an era of widespread misinformation and misconceptions about Islam, Yaqeen (يقين) is an Arabic platform dedicated to providing teenagers with accurate and well-researched, authentic responses to misconceptions about Islam. It serves as a trusted source for answering misconceptions in Islam based on authentic interpretations from the Quran, Sunnah, and trusted scholars. By addressing misconceptions with clear explanations and making Islamic knowledge easy to understand, Yaqeen helps teenagers develop a deeper and more confident connection with their faith. Our goal is to ensure accessibility to verified religious answers, strengthening their understanding and trust in Islam. The platform also includes a simple smart search system that supports keyword search, semantic search, and optional Fanar AI re-ranking and direct answers, making it easy for users to quickly find accurate answers from our curated Islamic dataset.
Project ID = S25SDP 14
Supervisor: Khaled Khan
Fagr Tahir, Kashaf Izhar, Maira Waheed
SkillUp is a mobile application designed to make skill-based learning more accessible and engaging, particularly for individuals in disadvantaged communities. Unlike traditional education platforms that focus on academic knowledge, SkillUp focuses on practical skills such as cooking, painting, photography and more. The app offers an inclusive environment by tailoring courses to meet the needs of a diverse range of individuals. For learners, SkillUp solves several common problems, including limited access to quality education, language barriers, and the lack of flexibility in traditional education. The app has a user-friendly design where users can explore and enrol in courses, track their learning progress, bookmark their favorite courses, and get personalized recommendations based on their interests or learning habits. SkillUp also supports instructors by giving them a dedicated space to design and manage their own courses. They can connect with students through interactive features and track student analytics. Most importantly, the platform allows instructors to earn money from their skills and share their knowledge with a larger audience, which is especially useful for skilled people in remote or low-income areas. Therefore, SkillUp is not just a learning tool. It is an application that promotes broader goals of access to online learning, skill-building, and financial growth.
Project ID = S25SDP 15
Supervisor: Amr Mohamed
Abdullah Khan, Ahmed Yousef, Mohammad Kasif, Mhd Hadi Nouh
Unmanned Aerial Vehicles (UAVs), commonly known as drones, are increasingly used in emergency response, crowd monitoring, and search-and-rescue (SAR) operations. A key challenge in these scenarios is detecting people in a specific area quickly and safely, especially in cluttered environments where traditional ground-based searches are slow, labor-intensive, and risky. Autonomous drones with onboard sensing and decision-making can offer a more efficient alternative. This project presents a low-cost multi-drone system architecture for detecting human presence using an RGB camera and LiDAR sensors. Each drone is based on a Holybro X500 V2 quadcopter with a Pixhawk autopilot and Raspberry Pi 4 companion computer. Instead of random exploration or complex learning behaviours, the drones follow optimized, pre-determined flight paths that can be divided among several vehicles to achieve systematic coverage of the search area. LiDAR-based obstacle avoidance maintains safe distances from surrounding objects, while an onboard object-detection model analyses the camera feed in real time and reports GPS coordinates of potential targets to the ground station. In this work, we implement and test a single drone as a representative unit of the swarm and demonstrate stable autonomous flight, reliable basic obstacle avoidance, and successful human detection in cluttered test environments, showing that the approach can be scaled to multiple cooperating drones for future SAR applications.
Project ID = S25SDP 16
Supervisor: Uvais Qidwai
Abdulaziz Al-Rumaihi, Nasser Aljufairi, Turki Al-fakhri
Fencing is a sport that requires exceptional reflexes, precision, and timing. One of the key challenges for athletes is maintaining and improving reaction time through regular practice. However, traditional training methods depend heavily on human coaches or partners, which may not always be accessible. This limitation has inspired the development of AlMubarez, a semi-interactive fencing manikin designed to simulate basic fencing movements using a mobile platform and a controllable robotic arm. The goal of this project is to build a functional prototype that allows fencers to practice their targeting, distance control, and movement coordination with a dynamic, programmable opponent. In this first phase, the system focuses on the mechanical and control aspects of the design. The manikin is mounted on a platform capable of forward and backward motion, simulating an advancing or retreating opponent. The arm features a wrist equipped with servo motors to mimic various fencing postures and angles. By emphasizing simplicity, modularity, and affordability, the AlMubarez system is designed to serve not just elite athletes but also fencing clubs, educational institutions, and individuals seeking practical solo training tools. The platform provides a foundation that can be easily expanded in future phases to include features such as reaction-time measurement, smart feedback, and data analytics. Overall, this project combines principles from mechanical design, control systems, and embedded programming to create a scalable and accessible solution for fencing training. It supports the growing interest in integrating technology with athletic development and demonstrates the potential of student-led innovation in real-world sports applications.
Project ID = S25SDP 17
Supervisor: Saleh Al-Hazbi
Mohammed Al-Kuwari, Saleh Bashraheel, Ali Al-Ghezaiz, Abdulaziz Alzahr, Saeed Alansari
This senior project presents an autonomous traffic violation detection system that leverages unmanned aerial vehicles (UAVs) and computer vision technologies to enhance road safety monitoring at Qatar University campus. The system addresses critical limitations of traditional fixed surveillance infrastructure by deploying a custom-built hexacopter drone equipped with high-resolution cameras and real-time video processing capabilities. The core detection pipeline utilizes YOLOv11 object detection models enhanced with an Adaptive Grid Learning Algorithm to identify multiple traffic violations including illegal parking, speeding, red-light violations, and unauthorized turns. The drone platform, constructed on an F450 frame with Cube Orange flight controller and Raspberry Pi 4 companion computer, provides autonomous navigation through pre-programmed routes while streaming video data for analysis. A PyQt5-based graphical user interface enables real-time monitoring, violation logging with GPS coordinates and timestamps, and drone control operations. Initial testing demonstrates successful integration of hardware components with MAVProxy flight control and accurate violation detection from aerial footage. The system offers a scalable, cost-effective solution for automated traffic enforcement that can adapt to diverse environments without extensive infrastructure modifications, making it particularly suitable for dynamic campus settings and expandable to broader urban applications.
Project ID = S25SDP 18
Supervisor: Mahmoud Barhamgi
Abdulmagid Al-Godimy, Abdulrahman Alshawabkeh, Ibrahim Aboulibdah, Faisal Al-malk
Parking congestion at Qatar University significantly impacts campus efficiency, resulting in wasted time, increased frustration among students, and elevated carbon emissions. Students frequently encounter difficulty locating available parking spaces, with disabled students facing even greater challenges due to limited accessibility. The absence of a real-time intelligent parking management system exacerbates these problems by causing delays and hindering smooth campus mobility. This project introduces QUPark, an innovative smart parking management system designed to resolve these issues effectively. Utilizing advanced AI-driven computer vision technology integrated with camera infrastructure, QUPark identifies available parking spots in real-time. The solution features a user-friendly smartphone application that provides immediate information about parking availability, directions to the desired building, EV charging station locations, and a convenient "Find My Vehicle" feature. To ensure inclusive usability, the app incorporates specific accessibility features tailored for users with disabilities. Additionally, robust cybersecurity measures, including AES-256 encryption and secure user authentication, safeguard sensitive user data. Key achievements of the project include the successful design of a scalable layered MVC system architecture integrating MongoDB, Raspberry Pi camera modules, and Google Maps API. Extensive market research confirmed significant demand for a smart parking solution, with over 80% of surveyed students experiencing regular parking issues and expressing interest in the application. Completed milestones comprise the systems high-level design, comprehensive UI/UX mockups, detailed database designs (ERD), defined use cases, and AI integration. The project also addressed ethical and legal standards related to user privacy and accessibility. To sum up, QUPark stands out as a scalable, secure, and inclusive parking management solution tailored specifically for Qatar University. It not only optimizes parking efficiency and reduces carbon emissions but also serves as a potential model for similar institutions aiming to enhance campus mobility and environmental sustainability.
Project ID = S25SDP 19
Supervisor: Saleh Al-Hazbi
Anandu Hirosh, Muhammed Ajmal Iqbal, Mohammed Faheem Ali Zaidi, Safwan Kavil
In the rapidly evolving digital environment of today, cybersecurity threats continue to increase in complexity as well as frequency. However, the tools that are available to counteract these threats remain fragmented, overly difficult to use and overly reliant on centralized infrastructures that are prone to being single points of failure. Many of these traditional cybersecurity tools are bloated with adware or lack transparency in their operations on top of requiring significant technical knowledge for their configurations, understanding and use. These gaps between usability, privacy and security become prime targets for the increasingly sophisticated attacks that are spreading about in today's digital world. This project addresses these issues by proposing a unified solution that combines advanced scanning features and threat detection tools with a decentralized anomaly detection and a dedicated compliance module, all tied together via a user-friendly and highly intuitive chatbot-powered user-interface that uses natural language to trivialize the use of advanced security tools. The proposed security solution, Adamas, is a next-generation cybersecurity application integrating federated learning with blockchain-based secure logging, real-time scanning and automated compliance assessment featured through an intuitive and completely local chatbot-based user-interface. The use of federated learning creates a privacy-preserving architecture that allows for the detection of anomalies without exposing sensitive user data, whilst the usage of blockchain ensures that recorded security logs remain tamper-proof and easily traceable. Moreover, the compliance assessment module gives the users an easy way to assess their compliance to security best practices and the chatbot powered interface trivializes the use of such advanced security features by utilizing natural language for increased accessibility for all user-types. The design features of the project system are further validated via a survey-based user requirements assessment which validates the direction of the project by highlighting the priorities of potential users. Furthermore, the class diagram, use-case diagram and chosen design patterns allow the system to be highly scalable for large-scale deployments. In conclusion, the project proposes a novel and privacy-conscious security solution that is tailored to modern threats and user expectations, with a focus on decentralized intelligence, secure logging and intuitive interactions.
Project ID = S25SDP 20
Supervisor: Rehab Duwairi
Mohammad Ahmad, M.Reyad Tuama Halabi, Obada Alrefai, Omar Qutb
Many high school graduates face uncertainty when deciding their future, especially when choosing a college major or whether to pursue higher education. This confusion often comes from a lack of guidance, external pressure, or not fully understanding their own interests. Making the wrong choice can lead to frustration, while the right one can set the foundation for success. Our project offers a smart and supportive mobile app that helps students make better academic decisions. Through a series of personalized questions, the app identifies their strengths and interests, then suggests majors that best match their profile. The goal is to help students feel more confident and aligned with their chosen path. The app also addresses key issues like the lack of academic advisors and scattered university information. By providing a 24/7 intelligent chatbot and centralizing important academic data, students get continuous guidance whenever they need it. Technically, the project combines cross-platform mobile development with elements from AI, user experience design, and educational psychology. This mix allows the app to deliver a personalized and accessible experience for all users. In short, our solution aims to guide students toward choices that reflect who they are, leading to a more rewarding and successful educational journey.
Project ID = S25SDP 21
Supervisor: Mohammad Saleh
Ali Zair, Muhammad Shehryar Qureshi, Omar Amin, Syed Subzwari
As phishing attacks have become a widespread nuisance and increasingly sophisticated, it is essential to protect users from them. Many internet users around the globe have fallen victim to email phishing attacks, bringing them financial distress, emotional anguish, or even wasted time. The issue is also observable in Qatar, with recipients receiving an exorbitant number of fraudulent emails, given Qatar's classification as a high-income country. Attackers leverage human weakness by impersonating figures of authority to pressure individuals to comply. These fraudulent messages seek to trick unsuspecting users into revealing sensitive information about themselves and/or making huge payments that harm users financially. Despite the growing awareness of phishing attacks, people may still struggle to recognize scam emails. Additionally, distinguishing legitimate emails from scams is getting more challenging as cybercriminals tailor their attacks to attain a higher success rate which threatens individuals and organizations alike. In fact, Experts at Kaspersky have voiced the need for robust email security solutions at the mail level to better protect corporations. To protect oneself, one must have sufficient knowledge in the field and look for telltale signs that give away whether the email is indeed a scam. However, not everyone has the know-how or expert knowledge to discern an attack from a legitimate email, making them a potentially vulnerable target. Our proposed solution for the project is an advanced email phishing detection and filtration tool that utilizes machine learning. By analyzing various aspects of emails, such as content patterns, and links, the add-on successfully identifies suspicious messages. In addition, the add-on provides explanations for its classifications, helping educate users and increasing awareness. Through these two main benefits, we aim to create a more digitally secure environment that evolves to phishing tactics, helps reduce financial loss, and saves users from emotional shock/wasting time. The project has advanced in terms of development. The system consists of separate frontend, backend GO API and python ML services that communicate via rest API's through docker compose. The ML model is an XGBoost classifier that achieved a final test accuracy of 97.85% by using SBERT vectors, surpassing the 96% threshold. Moreover, the model was trained on a dataset of 210,059 emails collected from multiple sources. Additional features were implemented, including VirusTotal scanning for URLs and attachments, explanation generation through an LLM wrapper, dashboard for viewing classification history and the ability to export scanned emails. This project demonstrates a combination of accurate phishing detection, clear user guidance and educational feedback. By integrating machine learning with a user-friendly interface, the system reduces the risk of financial and emotional harm from phishing attacks all the while supporting user education. The simple yet intuitive design enhances the emailing experience while making the add-in feel integrated into Outlook instead of getting in the way. Overall, the add-in represents a practical, scalable and effective solution to the ever-changing and complex problem of email phishing, contributing to a safer digital landscape for companies and users alike.
Project ID = S25SDP 22
Supervisor: Mohammad Saleh
Mohammed Ebrahim, Rashid Al-Bader, Sheikh Hasin Ishrak, Zubair Bin Jashim
Many students have useful skills in areas like programming, design, writing, music, or teaching, but they find it difficult to get real opportunities to use these skills. At the same time, there are many students who need help from skilled people whether for volunteering, small projects, learning something new, or even for paid work. However, because there is no proper or dedicated platform that connects students with each other for these purposes, many chances are missed. As a result, students' talents are not fully used, and people who need help struggle to find the right person. Currently, the student employment program at Qatar University only focuses on students who want to work within the university system. It is mainly designed for official part-time jobs and does not give support to students who want to use their skills in flexible ways like for short-term projects, peer learning, volunteering, or freelancing among students. This creates a gap where students either do not find help or cannot offer their skills easily. Solving this problem can help students connect better, work together on meaningful tasks, improve their personal and professional skills, and build a stronger support system among peers. Our project is an online platform designed to close this gap. It allows students to create profiles, list their skills, and describe what they can offer or what kind of help they are looking for. The platform includes features that allow students to search for others based on specific skills, needs, or interests. It also helps students with similar goals form small groups or teams. For example, a student who wants to start a business and needs a website can search for a programmer. A student who needs help with editing videos for a presentation can find someone who has experience in video editing. Even a student music group looking for a singer or guitarist can find someone with those skills. To make the platform more effective and trustworthy, we added useful features such as a skill-based search system, user profiles with detailed information, and a rating system so users can give feedback and build trust. We also included AI tools to suggest better matches between students based on their profiles and past activities. These features help save time and improve the user experience. Overall, this platform helps students showcase their hidden talents, find opportunities, and support each other in practical ways. It encourages collaboration, learning, and community building in the student environment.
Project ID = S25SDP 23
Supervisor: Mahmoud Barhamgi
Ahmad Chowdhury, Fahad Al-Eida, Hamad Al-Dosari, Hammad Muhammad Imtiaz
This project proposes the development of (ARHBO), a Tourism Guide Web Application tailored for Qatar, aimed at enhancing the travel experience by providing personalized, real-time information to tourists and local users. The primary problem addressed is the lack of localized and user-centric digital tools that showcase Qatar's hidden gems, cultural events, and authentic experiences, especially those overlooked by global platforms like TripAdvisor or Booking sites. Given Qatar's rapid growth as a travel and cultural destination, especially post-World Cup 2022, addressing this gap is essential to improving visitor engagement and leveraging technology to streamline travel planning within the country. The proposed solution is a web application featuring location-based listings, interactive maps, AI-powered recommendation engines, and user accounts with customizable dashboards. Users will be able to explore attractions, events, restaurants, and accommodations across Qatar through a seamless interface. The app also supports user-generated reviews and ratings, fostering a sense of community and shared exploration. This app addresses a genuine local need within Qatar's tourism ecosystem with technical depth and strong market awareness. It positions itself as a practical, innovative, and economically viable solution for Qatar's growing tourism industry.
More in Past Senior Projects
This project presents the design, implementation, and evaluation of a Secure Wireless Vehicle Access System that addresses vulnerabilities in Remote Keyless Entry (RKE) systems. While such systems offer convenience,...
Getting proper guidance at university can sometimes be frustrating as the traditional advising systems are not effective or ineffective. Many students face the challenge of not being able to get answers to question...
The desert climate in Qatar makes agriculture difficult due to water scarcity and extreme temperatures that affect the health of crops. The Smart Candy Farm is an innovative solution that addresses the biggest...