A project where I had to work by myself to design a completely functional system for a dental practice based in Sheffield. Given specific information related to the tasks secretary, hygienist and dentist had to perform I successfully implemented "Use Case Modelling", "Control Flow Modelling", "Information Modelling" and "Behavioural Modelling"

Such a fascinating essay about "Humanoid Robotics" and how their appearence might affect people's self-awarness and their empathy towards these machines. Another interesting subject researched in the essay was the Uncanny Valley effect, intensively studied by professor Masahiro Mori since 2009. According to this study people's empathy towards robots starts to decrease as they begin to look more like humans. Interesting question developed in this essay: "Are the robots going to become humans?"

In 3 weeks time me and a team of 4 students had to develop a completely functional application for the dental practice. The task involved basically continuing and further developing our earlier work from the previous project. The application was successfully implemented in Java using JDBC Connector and managed to register a new patient to the database, subscribe him to a new healtchare plan and book new non-overlapping appointments on a calendar for both partners.

Along with a team of 3 another students, we developed a full Cellular Automata genetic algorithm which fully simulated how the cells are reproduced in a living system, how they persist from one generation to another and how they die in the end. We also simulated John Conways' Game of Life and Forest of Fire by using von Neumann and Conway's neighbourhood rules to model the fire spread in a real forest. This image shows different generations of cells which were caught on camera to show how the fire was spreading from generation to generation.

A full Solitaire game developed in 2 stages: initially I had to code and implement all data structures used, all helper functions and an autoplay function which moved suitable cards from Reserves or Columns to Foundations. The second stage asked me to build a program which would play a game to completion and another one which would play 100 games to display the average score. I managed to play a game to completion (eODeal 72) where all the cards were moved to Foundations. On average the program managed to move 19-20/52 cards to final positions.

Optical Character Recognition System succesffully developed in Python. Given a scanned input page of a book for example, the program managed to recognize on average 99.93% of the characters. It performed 20 different tests on various test pages. Its performance varied from 74% correctly recognized and classified characters to 99.93%, depending on the noise from test pages. The image on the left shows the best score gained by my classifier (test2) and the one gained on the noisier page (test1).

A fully functional website entirely developed for an important client of University of Sheffield who was actively managing multiple water research projects. Having complete, but yet quite complex specifications, the client's representative asked me and a team of another four students to build a web platform which could gracefully help them manage their research projects database. You can see on the canvas on the left the admin area, a project's page, the news and the main page containing all the projects from the database, along with the poster for this project.

Using the Mindstorm EV3 Robots provided by the faculty, I had to work in a team of two to basically program the little robot to behave as Braitenberg vehicles 1 and 2. Our robot was simulating 4 feelings described by Braitenber as 'Love', 'Aggression', 'Fear' and 'Explorer'. By using 2 ultrasonic sensors and a light sensor we successfully achieved all the 4 'feelings'. If the robot 'Loved' the object detected, it kept following the object until getting very close to that. Click the image on the left to check up 'Love'!

A one week real life engineering challenge where together with a team of another five engineers, I had to provide a solution to help in early detection of eye diseases. Me and my team proposed the winning solution of a simple retina scanner gun, which after shooting the tested eye, it could be placed on a charging deck and download the images taken into a computer. There, by using a complex classifier, a score of how healthy the picture of eye looks might be computed in under 2 minutes and the person in cause could then be redirected to a specialised consult in case of any anormalities.