Supervisor |
Title |
Brief Description |
Pre-requisites |
Type |
Availability |
M. Junaid Anjum |
Edge-Enabled UAVs for Autonomous Flight and Air Quality Forecasting |
This project develops an Edge AI–powered framework that enables UAVs to perform autonomous flight operations while simultaneously predicting Air Quality Index (AQI) levels. By leveraging machine learning models deployed on edge devices, the system reduces latency, supports real-time decision-making, and enhances environmental monitoring efficiency. |
Machine Learning and Proficiency in Python |
Research + Development |
Yes |
M. Junaid Anjum |
Agent-Based Decision Support for Course Registration and Academic Planning |
This project designs a smart academic registration system built on multi-agent architecture. Each agent is responsible for tasks such as verifying student records, checking course prerequisites, managing seat availability, and resolving scheduling conflicts. The result is a more automated, adaptive, and efficient course registration experience compared to traditional systems. |
Machine Learning and Proficiency in Python |
Mobile Application Development |
Yes |
M. Junaid Anjum |
Collaborative FL-Based Early Warning System for Floods Detection |
This project applies Federated Learning (FL) to develop a privacy-preserving early warning system for flood prediction. Distributed data sources such as IoT water-level sensors, rainfall gauges, UAVs, and satellite imagery contribute to model training without sharing raw data. The aggregated global model provides accurate, real-time flood risk alerts, supporting timely evacuation and disaster management. |
Machine Learning and Proficiency in Python |
Research + Development |
Yes |
Dr. M. Tayyab Chaudhry |
Crypto Analytics |
To apply data analytics over various crypto data. |
Mobile and Web Application Development |
Mobile and Web Application Development |
Yes |
Dr. M. Tayyab Chaudhry |
BlockChain Based E-Chalan System |
Development of e-chalan system for traffic ticketing. |
Mobile and Web Application Development |
Mobile and Web Application Development |
Yes |
Dr. Tahir Maqsood |
Tutor Finder |
Tutor finder will be a web and android based application that will allow students and parents to find teachers in their vicinity. The system will enable teachers to register themselves in the system through web-based interface or a mobile application. The teacher will enter its profile information and portfolio along with the subject of expertise. The students and parents will be able to search for tutors for a course or topic. The main functionality lies in providing feedback for tutors by students/parents. Feedback will be provided both in terms of text and ratings. Based on comments/ ratings the parents will be able to select the tutor. Moreover, in addition to subject based expertise, topic-based expertise and feedback can also be incorporated. The application will allow students/teachers to schedule the tuition time based on available slots. |
Mobile and Web Application Development |
Mobile and Web Application Development |
Yes |
Dr. Tahir Maqsood |
Travel Buddy |
The Travel Buddy is a web and mobile based application that will help travellers and tourists to get the best accommodation while they are travelling. Through the application, travellers can search and reserve a room(s) in the visiting city or tourist place according to their preferences and requirements. This will also help the house owners to rent out their vacant rooms. Moreover, Travel Buddy will allow users to join a tour or users can also plan their own tour, which others can join. |
Mobile and Web Application Development |
Mobile and Web Application Development |
Yes |
Iqra Obaid |
WattWarrior: Gamify Your Way to Energy Conservation |
The objective of this project is to develop an innovative energy conservation program aimed at encouraging individuals to take responsibility for their energy consumption. The program will utilize gamification principles and collaborate with utility companies to provide households with personalized energy consumption data, neighbor comparisons, and milestones tracking. The goal is to inspire people to reduce their energy usage. |
Programming |
Android based Application |
Yes |
Dr. Junaid Akram |
Building a benchmark (dataset) of vulnerable source code
+
Analysis of vulnerable repair patterns |
Vulnerability detection and creation of patch files have improved the ability of fast overcoming the security threat in software systems. An automatic exploit is possible through patch files. Patches to vulnerable code are very important to make sure security measures. Whenever a patch is released, an attacker (Hacker) can easily use this patch to find exact location of this buggy code in a very short time and can automatically create an exploit.
Building a Benchmark (dataset) by extracting the vulnerable code from the patch files at different levels of granularities (function-level, file-level, component-level) that will help the security people in better understanding of vulnerabilities structure. This benchmark will also help us to overcome cyber security attacks by removing similar vulnerable code fragments from other software. The detection of similar vulnerable code fragments can be implemented by using any code clone detection technique.
+
Analyse the vulnerability patches to explore the most often changes applied in the code to remove the vulnerability, and recommend the lesson learned to the community and the developers, so that they can minimize those mistakes that are causing vulnerabilities. |
Interest in Software Security
&
Strong Programming Skills |
Research & Development |
Yes |
Dr. Junaid Akram |
Detection and analysis of IntraProject vs InterProject software dependency ratio |
Code clones in source code always bring a lot of troubles in software security maintenance. Previous research shows that almost 7% – 23% of source code is cloned in large systems. Even recent work shows that 22.3% of Linux code has been cloned.
In this project, we aim to analyse the source code dependency ratio between similar types of software, as well as in different versions of the same software. This work will help us to properly understand software (code) dependency. Case study (Linux Versions) |
Interest in MIS
&
Strong Programming Skills |
Research & Development |
Yes |
Dr. Junaid Akram |
Auto Patching: Applying patches to the similar code fragments throughout the code repository |
A single exploited software vulnerability can cause many damages to an organization. Annual losses caused by cyber-attacks have been reported as $226 billion through world-wide (report in 2020).
In this project, we aim to develop a tool for automatic patching mechanism (replace the vulnerable files). This tool would be helpful for the organizations that provide open-source facilities (i.e., GitHub). By using our tool, it would be easier to update (apply patch) the code in all the applications that share the similar code.
For example, if several software on GitHub have used/copied Linux-kernel code, and a vulnerability has been discovered in the kernel, although this vulnerability will be removed by updating this vulnerable code through a patch file in the latest version. But it (vulnerability) remains in other software by not knowing that a patch has been released. Our tool will help to apply this patch to all other similar applications that have shared the similar code. |
Interest in Software Security
&
Strong Programming Skills |
Research & Development |
Yes |
Dr. Junaid Akram |
Software vulnerability prediction model to predict existence of vulnerabilities in the software |
The existing software systems have become larger and more complex. Software vulnerabilities have also increased rapidly, potentially causing an increasing number of serious security threats. These vulnerabilities should be removed from the systems as soon as possible.
In this project, we aim to detect vulnerabilities from the system by exploring patch files. Based on repair pattern in these patch files, we will train a model on these patterns and then predict the nearly or exact vulnerable code fragment in the open-source software. The models aim to automatically learn software vulnerability knowledge and patterns from the vulnerability dataset. The purpose of these models is not to identify vulnerabilities but also to predict the number of vulnerabilities that are likely to be appeared in the code. |
Software Security +
Machine Learning +
Strong Programming Skills |
Research & Development |
Yes |
Dr. Junaid Akram |
Automatic test data generation for data processing systems by using different mutation operators |
The ability to properly generate test data is often necessary, and prerequisite for automated software testing. The test data can be of two types; valid data or faulty data. It’s important to test system’s behaviour for both types of data. Testing the correct behaviour of the systems in the presence of faulty data is extremely expensive, especially if the system has to be tested against huge number of test inputs, and it all have to be done manually.
This project aims to develop an automatic approach to generate test data for complex systems. The tool will generate faulty test data as well as valid test data to explore the input space of the SUT (System Under Test). For generating the valid data, to be fit for its intended purpose, the data usually has to satisfy various logical constraints between the data inputs. |
Interest in Software Testing
&
Strong Programming Skills |
Research & Development |
Yes |
Imran Latif |
Weather Forecast Website |
Building a weather forecast website that provides you the information about the weather conditions based on your search locations.
During this project, you will build features like weather search via zip code, view current temperature and humidity, check wind speeds, and another forecast for the next 5/7 days. |
Preferably if the students had studied the Web Development course |
Development |
No |
Samia Arshad |
Email notification system on WhatsApp with content filtering |
The objective of this project is to develop an intelligent WhatsApp notification system that notify when an email is received. Moreover, the system will generate an appropriate response based on the content of email. The response will be sent to the sender of the email as an auto-reply, acknowledging the receipt of the email. This project can be useful for business and individuals who receive a large volume of emails daily and want to automate the process of filtering and responding to emails by using WhatsApp. |
Programming skills |
Development |
No |
Samia Arshad |
Building an AI powered chatbot using open AI’s GPT-3 |
The objective of this project is to build a chatbot using Open AI's GPT-3. The chatbot should be able to understand and generate text in natural language and carry out a conversation with users on various topics. The chatbot will be trained on a diverse range of topics and should be able to answer questions, provide recommendations, and engage in small talk. |
AI, machine learning |
Research and development |
No |
Dr. Tahir Maqsood |
Parental Control Application for Mobile |
The objective is to create an application that will enable parents to control and track all the activities being performed on the mobile phone. This will be an android based application that will allow parents to control the application usage, i.e., which application can be used at what time. Moreover, game and internet control functionalities will be provided. Similarly Call and SMS restriction functionalities will also be provided. In application/game control, parents can specify that which games can be played as well as restrict the game timings. In internet control usage the application will control that which sites can be visited as well as limit internet usage time. This application stores all previous record of game history as well as browsing history. The logs are uploaded to the server and can also be emailed to the parents. Parents can also view the activity log through a web-based interface. The app will provide remote settings and monitoring of mobile usage to parents. |
Programming |
Android based application |
Yes |
Dr. Tahir Maqsood |
Online 12-piece (12-Goti) Game |
The project 12 Piece (Goti) Game is an android based app. It’s a two-player game, in which each player has 12 pieces (Goti). These pieces can move in different directions on the valid position of the court. If a player, crosses a piece of opponent player, then the player will achieve 1 point and that particular piece of opponent is removed from the court. In this way, whoever manages to achieve 12 points will be the winner. This game is a strategy board game for two players. The Game board comprises a 5x5 two-dimensional matrix.
The App will allow two players to play the game while connected through the LAN or the Internet. |
Programming |
Game Development |
Yes |
Dr. Adnan Ahmad |
Smart Internet censorship |
Internet censorship is a way in which an authority has its control over the user’s surfing over the web. It includes the control of what can be seen viewed, accessed or published over the web. Internet censorship is applied over the web traffic due to some religious, business or moral reasons. Today internet surfing is becoming an issue which requires great attention of ISPs in almost every part of the world and they are trying to solve these issues by different ways. Some websites are not required in the community and the presence of such websites/blogs etc. is harmful for the users due to some cultural and/or political reasons. This final year project considers the very problem and proposes a smart internet blocking system based on the contents present on a website. For further details please visit Dr. Adnan Ahmad in Faculty Block Room 37 |
software engineering, algorithms, good programming and analytical skills |
Research & Development |
Yes |
Dr. Adnan Ahmad |
Sandboxing |
Sandboxing refers to the security mechanism which is used to separate various running programs in the same system. It is normally used to execute untrusted code, so that if there is any harm in its execution, it remains limited to an isolated environment only. Let us consider the classic dilemma, where a person has received a file and eager to see what is inside, but at the same time afraid of infecting his system through a virus. One possible solution is to open the file in a controlled environment (sandbox), so even if it is infected, the effect can be localized. Without a proper sandbox mechanism a system always faces threats of viruses and Trojans every time a code executes. This final year project deals with the development of a sandboxing environment and helps students to deeply understand programming fundamentals, operating systems and access rights. For further details please visit Dr. Adnan Ahmad in Faculty Block Room 37. |
|
Research & Development |
Yes |
Dr. Rao Muhammad Adeel Nawab |
Identification of Proper Citations |
Probably one of the most interesting topics for future research on plagiarism detection is the automatic analysis of citations and references, a problem that, although mentioned already by Maurer et al. (2006), still remains unsolved. When a text fragment is properly re-used, i.e., including its corresponding citation or reference, it does not compose a case of plagiarism, and it would not be considered as suspicion. The development of techniques for discriminating between plagiarism and properly cited re-use requires, firstly, obtaining (or generating) diverse cases of quotations and, more general, proper and wrong citations. Afterwards, models that recognise proper citation patterns could be designed. |
Python, Java or Perl programming skill |
Research & Development |
No |