University of Malta

2010 Projects
UOM Main Page
Apply - Admissions 2016
Campus Map button

Final Year Projects Academic Year 2009/2010



Student: Anabel Bonello
Supervisor: Prof. Carmel Pulé


This work introduces a deeper illustration of the mathematical operators that may be used in multi dimensional filters or recognition systems, and provides the resultant waveforms, shapes and contours generated in various impulse functions that can select various parameters. Such concepts, in the areas of signal recognition and filtering would grow to produce observers as comparators which would operate continuously and supervise all activities and thus achieve corrective measures before disasters occur.
Project Objectives
 • Mathematical analysis of the behaviour of systems for multidimensional signals.
 • A General Recognition Function describing a Polarised Filter system.
 • Interpretation of the polarised Laplace Function and the Convolution Integral.
 • Operation on the Impulse Function to create a more general recognition kernel.
Project Methodologies
During this Project a fascinating world of space and time vector arrangement portrayed a marvellous understanding of the behaviour of signals. The following steps were carried out during the implementation of the project:
• Literature review of various analog and digital filters. Furthermore conventional results for the time and frequency responses through simulation were obtained and some characteristics and performances were observed through implementation.
• Analysing any system behaviour using important mathematical techniques, using the Laplace Transform and the Convolution Integral performed on multi dimensional signals.
• Presented an overall general recognition filter expression where it described the impulse function of a system as it selected space, position, timings, frequencies, and shapes. Futhermore polarisation in the space domain was performed on the 3 dimensional mathematical techniques.
Results and Achievements
By knowing how to synthesis, the character of the system, many forms may be obtained such as the complete impulse function of vertical dipoles, lenses and optical filters. One may look upon the system kernel and the software/ideal kernel used in Laplace and Convolution Integrals to be as matrix, composed of individual elements that operate individually. These matrixes are located on a more complex non-conventional s-plane and by applying these mathematical techniques through comparing the real and software kernels, errors are found. The complex multiplication of these matrix components would produce products with and without a meaning, but it is predicted that the dot product of components of the matrix multiplication would have a useful meaning in signal and scenario recognition.



Student: Michaela Camilleri
Supervisor: Prof. Carmel Pulé


The field of autonomous robotics has various applications ranging from game simulations to health and safety automations whereby such a system would be developed in order to locate the shortest route for a fire rescue operation in a building or the location of the shortest escape route during a mine explosion. As the name of this project suggests, a maze solving robot is a robot that can travel through a maze by changing its direction whenever any obstacle occurs in a particular direction. Thus such concept may be used to explore any particular hidden region without any human intervention.
Project Objectives
This project is intended to obtain experience in more than one area, including practical electronics handling information, power circuits, control systems and software engineering, which also includes a degree of activity which would be regarded as intelligence through a learning process. A robot is required to enter an unknown maze virtually, learn to find the closed and the open paths, and decide to select on a way out of the maze to freedom. Learning through errors is permitted through the first trial and while the duration of the learning activity is monitored to see how quickly it learns, then the second trial would be covered directly without any errors. This information shall then be transferred onto the hardware version of the robot so that the maze is also solved in its hardware version.
Project Methodologies
The following is a list of the main tasks involved in this project:
1. Conduct research on the following topics:
    • Maze creation and solving techniques
    • Stability control
    • Selection of an evaluation board
    • The sensing devices available
    • Robot building techniques (including the
       choice of the building components)
2. Design and development of a simulation in software for maze creation and maze solving.
3. Familiarisation with the evaluation board’s features.
4.Robot hardware design and assembly – both electronic and mechanical aspects.
5.Robot software design and development.
6.Testing and improvements of the system.
Results and Achievements
The maze simulation software was implemented successfully. Figure 1 shows a 20x20 maze and its solution. Furthermore a hardware version of the robot was built as can be seen in Figure 2. The route taken by the virtual robot was then transferred onto the hardware version of the robot and this worked successfully.



Student: Stephen Grixti
Supervisor: Prof. Ing. David Zammit Mangion


Inertially stabilized platforms (ISPs) are used to stabilize a broad array of sensors, cameras, telescopes and weapon systems. Although requirements for ISPs vary widely depending on the application, they all have a common goal, which is to hold or control the line of sight of one object relative to another object or inertial space. The most widely known application of all is the use of inertially stabilized camera platforms mounted on moving vehicles such as aircraft [1].
Project Objectives
The objective of the project is the design of a single-axis stabilized platform for airborne applications and is the first departmental step towards achieving a stabilization system for a small unmanned aerial vehicle (UAV). The system should be able of isolating a light-weight payload from any pitch rotations undergone by the host aircraft.
Project Methodologies
The basic principle throughout the project was to sense the pitching angle using appropriate sensors for airborne applications and applying a counteracting torque to stabilize the payload. The project was divided into logical phases so as to ensure correct implementation at each stage:
• Literature review and research of stabilized platform applications and technologies.
• Experimental setup of the gimbal-like structure and selection of components such as the driver motor and    Microelectromechanical systems (MEMS) sensors.
• System modelling in Simulink® and through other analytical techniques.
• Implementation of the stabilization system using a potentiometer to sense aircraft pitch attitude. Pitch sensing using appropriate MEMS sensors, evaluation of results and conclusions.
Results and Achievements
Satisfactory stabilization was achieved by sensing pitching attitude using a potentiometer. Formal testing and demonstration involved the use of a video camera mounted on the platform: camera line-of-sight along the pitch axis was correctly stabilized by the control system. The system is now to be further implemented and finetuned using the more appropriate MEMS sensors.
[1] J.M. Hilkert, “Inertially stabilized platform technology, concepts and principles”, IEEE Control Systems Magazine, pp. 26-46, Feb.2008.
[2] M.K. Masten, “Inertially stabilized platforms for optical imaging systems, tracking targets with mobile sensors”, IEEE Control Systems Magazine, pp. 47-64, Feb. 2008.



Student: Matthew Sammut
Supervisor: Ing. Marc Anthony Azzopardi

Research is currently being carried out with regards to the design of advanced cockpit display systems, capable of presenting large amounts of data in a very short period of time. This type of cockpit display could be implemented as a Head Mounted Display (HMD) and would allow the pilot to scan a display much larger than his field of view by simply moving his head. For this to be possible the HMD has to allow the aircraft to calculate the position and orientation of the pilot’s head within the cockpit. A previous project [1] involving the tracking of the orientation of the pilot’s head has already been carried out however a system which tracks the absolute position is yet to be implemented. Such research is relatively new and thus allows for several innovative implementations.

Project Objectives
The objectives of this project are to design an ultrasonic absolute positional head tracking system which: processes data independent of previous knowledge so as to avoid any drift and abbe errors, is relatively cheap yet reliable and accurate with a fast update rate, is light weight and not a nuisance to the user and is able to cover a reasonable area so as not to restrict the user’s movement.

Project Methodologies
The project deliverables were modularized into the following steps:
• Conducted a literature survey of di!erent types of positional tracking systems where di!erent systems were compared for their di!erent pros and cons until a most suitable tracking system was decided
• After it was decided that an ultrasonic tracking system best "ts the criteria, further research was carried out to determine a suitable algorithm to be used for the determination of the position of the user’s head.
• Development of an ultrasonic head tracking system which involved the determination of the main subsystems required within the system, design of circuitry with the help of an electronic simulator, implementing and testing of prototype circuits, implementing the required code in microcontrollers and interfacing electronic hardware with microcontrollers.
• Production and analysis of results with varying parameters. This involves observing the accuracy of the computed position of the user’s head and also analyzing the variation of the update rate of the system for di!erent conditions

Results and Achievements
Results include the analysis of the accuracy of the pilot’s head position: while static, during movement and when subject to di!erent temperatures. In addition to this the update rate is also to be analyzed at di!erent conditions. Simulation results have shown that a satisfactory accuracy can be achieved. Future work is suggested involving the data fusion of the absolute position of the pilot’s head together with its orientation in order to achieve a more accurate result.

[1] Fonk K., “Inertial measurement in light-weight head-mounted display systems”, Unpublished, B.Eng. Dissertation, Department of Electronic Systems Engineering, University of Malta, Malta, 2009.

Last Updated: 5 November 2012

Log In back to UoM Homepage