http://hdl.handle.net/123456789/13176 Thu, 16 Jul 2026 00:01:10 GMT 2026-07-16T00:01:10Z http://hdl.handle.net/123456789/20607 AIOT Based Vehicle Maintenance System Abdur Rehman, 01-135212-008; Zimal Faisal, 01-135212-102 The AIoT-Based Vehicle Maintenance System aims at enhancing the management of vehicles maintenance through the integration of smart tech using the focus on the user experience. User data on all aspects of part maintenance is available by means of a personal dashboard getting easy access to the vehicle’s condition. The system does more than just monitor; Real time alerts, part comparisons and condition specific predictive updates are also included in this system’s offerings. The system maximizes both cost effectiveness and overall reliabilty of vehicles. The platform provides personalized notifications as to when maintenance needs to be executed by examining weekly distance and average speed. App users are informed of specific instances when replacement of these particular components is due, and based on the user’s driving patterns they are given pertinent repair advice. Conditions of the roads are also analyzed to determine when the tasks of road maintenance should be carried out accordingly. Furthermore, the platform also evaluates drivers’ driving behaviour, by for example recording instances of when there are problems in the system and when there aren’t, so as to aid estimation of the life cycle of vehicle components. The workshops can take advantage of a web portal that comes with AI-based spare part suggestions, real time inventory tracking of a nearby inventory and comparative pricing alternatives and are given authority to send maintenance alerts to users customized to their full maintenance logs. A device with AIoT technology is also part of the system, tracking the vehicle’s travels with IoT sensors and sending the data straight to the mobile app so maintenance predictions are accurate. Also, the built-in scraper gathers real-time price data from Ali Express, so users can decide which items are best for them. The application is based on a advanced and scalability-friendly tech platform. The mobile aspect is created with Flutter, which will guarantee a natural and unified user experience. React.js is used to build workshops’ web portal to provide quick and interactive experiences. The Flask APIs handle the backend and data is stored in real time with Firebase. Training models requires Python and TensorFlow.js and Scikit-learn are added to offer important prediction capabilities. Using Git and GitHub, developers can control the code and make it easier to develop using VS Code, encouraging effective coding and easy collaboration within the team. Supervised by Ms. Iqra Javed Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/20607 2025-01-01T00:00:00Z http://hdl.handle.net/123456789/20606 Personal AI Educator: Learn Your Way M. Ihtesham Akram Awan, 01-135212-035; Umair Shahid, 01-135212-035 The rise of artificial intelligence and web technologies has reshaped how educational content is delivered and consumed. In response to the growing need for personalized and intelligent online learning experiences, this project introduces Personal AI Educator: Learn Your Way — a full-stack web application designed to offer adaptive, AI-assisted education tailored to individual learners. The platform leverages modern web technologies, including React.js for the frontend, NestJS for backend service orchestration, and PostgreSQL for persistent data storage. At the core of the system lies an integration with the DeepSeek API, which dynamically generates subject-specific lecture roadmaps, full lecture content, and corresponding quizzes. This enables each user to follow a structured, AI-curated learning journey based on their chosen subject. Lectures unlock progressively, with the user required to pass each quiz with a minimum score, ensuring active engagement and knowledge retention. The application begins with a responsive landing page that introduces the platform’s purpose and highlights available courses. It offers both traditional email-based and Gmail login via Firebase Authentication. Once signed in, users are directed to a subject selection interface. Each subject opens access to three interactive modules: “Learn with AI,” “Nearby Institutes,” and “Available Courses.” The Learn with AI module provides AI-generated educational content and quizzes. The Nearby Institutes module integrates Google Maps to display location-based institutes offering related subjects, while the Available Courses section showcases scraped online courses from platforms like Coursera and edX, offering users external learning opportunities. User data, including quiz scores, lecture progress, and saved content, is securely stored in PostgreSQL, supporting continuity across sessions. The system follows a modular, test-driven development approach, with components evaluated through unit, integration, and end-to-end testing strategies. By combining artificial intelligence with scalable web infrastructure, Personal AI Educator redefines how learners interact with online education. It not only delivers personalized content but also enriches the learning experience through real-time content generation, structured progression, and curated academic resources — making self-paced education more intelligent, accessible, and impactful. Supervised by Mr. Adnan Jelani Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/20606 2025-01-01T00:00:00Z http://hdl.handle.net/123456789/20603 Social Media (Twitter) Bot Detection M. Muneeb Ahmed Kiani, 01-135212-063; Hamid Nisar, 01-135212-042 The Social Media Bot Detection System (SMBDS) is an innovative solution designed to address the growing prevalence of bot-driven activities on Twitter, a platform central to global communication and information sharing. This system provides an integrated approach to identifying, analyzing, and managing suspicious accounts, ensuring the authenticity and integrity of interactions. By replacing traditional manual monitoring methods, SMBDS facilitates real-time data sharing and automated detection of bot-like behaviors, enhancing efficiency, accuracy, and transparency. Built on [state the technology platform, e.g., Python, etc.], the system employs advanced data scraping, machine learning algorithms, and behavioral analysis tailored to Twitter’s unique ecosystem to detect anomalies such as spamming and repetitive posting. With minimal human intervention, administrators can easily flag, unflag, and update the statuses of suspicious accounts using an intuitive interface, ensuring rapid and informed responses to potential threats. Access to real-time analysis enables better decision-making, reducing manual errors and ensuring effective resource allocation. This project addresses critical challenges, including slow manual detection, lack of transparency, and the inefficiencies of traditional monitoring, by automating the identification and mitigation of malicious bot activities. The Social Media Bot Detection System offers Twitter a robust tool for improving platform safety, preserving user trust, and fostering an authentic and secure environment for communication. Supervised by Dr. Saba Mahmood Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/20603 2025-01-01T00:00:00Z http://hdl.handle.net/123456789/20604 Smart Waste Bin System Hina Khurshid, 01-135212-032; Syed Muhammad Ahmed Us Subhan, 01-135211-081 Waste management is the major issue faced by both developed as well as developing countries like Pakistan. Considering waste issues in Pakistan, an effective and costefficient smart waste bin system is proposed. The objectives were to develop a smart waste bin system, its application to the field and integration of GIS for develmaps. The smart waste bin system is based on two ultrasonic and gas sensors for the central micro controller linked to waste level and gas detection that provides a link between the functioning of the components physically as well as through programming. A GPS module was used for location detection and GSM module for SIM connectivity. Thingspeak was selected as a cloud that collects data per minute from sensors. The Thingspeak data was based on location latitude and longitude, gas detection and waste level. This data is shown in the form of graphs. The "GIS Integration Smart Waste Bin System" is an innovative solution designed to enhance waste management efficiency through integration of Geographic Information Systems (GIS) and IoT technology. This project aims to address the challenges of urban waste collection by utilizing smart waste bins equipped with sensors to monitor fill levels in real-time. Data collected is transmitted to a centralized system, enabling waste management authorities to optimize collection routes and schedules based on actual bin usage. By incorporating GIS, the system visually maps waste bin location facilitating better planning and resource allocation. Users can access this data via a user-friendly application, promoting community engagement and encouraging responsible waste disposal. Supervised by Mr. Ahtesham Noor Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/20604 2025-01-01T00:00:00Z