Dissertation Project Ideas
 
A list of project ideas that require some further thoughts and discussions.
 
  • Improving "code2uml", a tool for generating UML class diagrams from java .class and .jar files. See here for more.
  • Replicating James Lovelock's Daisy World (A Self-Organising Biosphere). See here for more.
  • Improving AI Decision Modelling through Utility Theory. See here for more.
  • Fuzzy Logic for Social Simulation: Using Juzzy from within AnyLogic. See here for more.
  • Develop Your Own Simulators for Discrete-Event Systems from scratch. See here for more.
  • Drivers' En-Route Divergence Behavior Modeling Using the E-BDI Framework. See here for more.
  • Agents Interacting on GIS Maps. See here for more.
  • Android App for Modelling the Dynamics of "My Citations" in Google Scholar
  • Building a Computer Game with Advanced Fuzzy Logic Decision Making Embedded.
 
 
A Choice of Previous Dissertation Projects (use these for inspiration)
 
The Digital Occupant
 
Monitoring the uptake of new equipment in private homes has often some ethical issues (intrusion of privacy) and is also quite cost intensive. We were interested in exploring the opportunities of using digital occupant for such tests. The digital occupant (individual / group) could be used for testing and calibrating newly developed devices (e.g. smart meters) that are part of the house of the future. This way of testing is non-intrusive and avoids the ethical issues that people might have when doing research with humans in the loop. In this project, the student's job was to explore existing solutions and then to build his own digital occupant and test environment using agent-based modelling and simulation.
 
 
Improving Road Safety for Lorry Drivers through Data Analytics and Agent-Based Simulation
This project was jointly supervised by me, Dr Grazziela Figueredo and Professor Bob John.
 
The student's project was linked to a bigger research project that is aiming to understand and improve road safety for lorry drivers around London through data analytics and agent-based simulation. Lorry data regarding vehicle tracking, telematics and weather conditions have been gathered throughout a period of three months. The collected information includes incidents such as lorry harsh cornering, harsh braking and speeding. In this project the student's job was to fuse and analyse the data and determine those areas in London of high danger for drivers and their causes. He was also asked to come up with stereotypes of lorry drivers based on the available data.
 
 
The Elephant Welfare Project
 
The idea for this project came from a discussion I had with Lisa Yon from the Vet School. She is interested in studying attraction to or repulsion from physical and social features of the enclosure for zoo elephants, with a view to identifying those elements of their environment which are particularly important to the elephants. In order to study the complexity of social and physical interactions between elephants and with their environment I proposed to employ simulation. A simulation method often used when the focus is on modelling individuals and their interactions is agent-based modelling and simulation. There are different ways of defining such agent-based models and each research field has its preferred way of modelling. Behavioural animal models are usually formulated using equations that are based on theories and qualitative observations and empirical data is used to validate these models. In this project, the student's job was to test a new approach using Finite State Machine diagrams together with social theories to define the elephant agents and their behaviour.
 
 
Energy-Efficient Data Forwarding in Mobile Wireless Sensor Networks
This project was jointly supervised by me and Dr Sameh Zakhary.
 
A typical Mobile Wireless Sensor Networks (MWSN) consists of a number of mobile nodes, and each node represents a small device (sensor) that has limited computing and storage capabilities. These sensors are capable of measuring their environment and reporting back various events (e.g. temperature, pressure, GPS location and humidity) to one or more base station(s) that can either be static or mobile. Many protocols have been developed to enable nodes to become autonomous acting as an independent agent to transmit their own data, as well as to collaborate with each other to move data around the network. As sensors have limited energy and storage capacity, each sensor aims to get as many events reported back to the base station as possible with the least cost. Also, sensors maintain collaborating with each other to prolong the lifetime of the whole network. In this project, the student's job was to develop an agent-based model for forwarding information between nodes to facilitate communication in MWSN, and compare its performance to existing solutions such as epidemic routing. As a case study we used an animal movement monitoring scenario, where we wanted to test how quickly the two different forwarding protocols allow a veterinary to respond to stress signals sent from sensors mounted on the animals (e.g., when one animal gets ill or insured) and how to best to schedule the doctor's visits to do routine health checks on animals.
 
 
PhD Students: Main Supervisor
 
  • James Hey
    • Urban Social Simulation in the Context of Sustainability

  • Kwabena Amponsah
    • Distributed Computing and Optimisation of Complex Urban Systems

  • Jonathan Thaler
    • Functional Agent-Based Modelling and Simulation

  • Sudhir Venkatesan
    • A Framework for Multi-Method Simulation of Antiviral Treatment for A(H1N1) Pandemic Influenza

  • Olusola Faboya
    • The Use of Object Oriented Agent-Based Social Simulation to Stimulate Transport Mode Shift of People

 
 
PhD Students: Additional Supervisor
 
  • Alexandra Cristina Dediu
    • Office space allocation problem: human computation and simulation approach

  • Shadab Mashuk
    • Application of indoor positioning to understand building energy consumption using an occupancy model

  • Tim Whiteley
    • Integrated whole system modelling and optimisation of city resource flows

  • Kunpeng Wang (writing up)
    • Multi-scale model integration for the large-scale analysis of complex urban energy system

  • James Burnett (writing up)
    • What can user data relating to proximity and orientation tell us about real-world vs. simulation for interactive content delivery

  • Felix Osebor (writing up)
    • Sustainable urban mobility: A modelling framework for cities in rapidly developing countries

  • Xia Li (writing up)
    • Port operation evaluation with simulation and fuzzy based multi-stakeholder multi-criteria decision analysis

 
 
PhD Students: Advisor
 
  • Ashley Lewis
    • Interdisciplinary approaches to assessing urban sustainability: An ethnographic study of interdisciplinarity in practice.

 
 
PhD Completions
 
  • Dr Tuong Vu
    • A Software Engineering Approach for Agent-Based Modelling of Public Goods Games

  • Dr Jacob Chapman
    • Multi-agent stochastic simulation of occupants' comfort and behaviour

  • Dr Mazlina Abdul Majid
    • Human behaviour modelling: An investigation using traditional discrete event and combined discrete event and agent-based simulation
This site uses cookies to anonymously measure how people use it!