360 Virtual Tour of GCU

Overview: With COVID-19 outbreak, Maria’s PhD supervisor Professor Gordon Morison decided to purchase a 360 camera and for a short period repurpose Maria’s Data Visualization talents. This was initially for the Department of Computing, but immediately became School then subsequently University wide.

Challenge: Design and development of a 360 Virtual Tour website so that future students can navigate through Glasgow Campus facilities and have an immersive campus experience.

Scope: Six weeks. From research to conception, prototyping, testing and launch.

Tools: HTML, JS, CSS, Marzipano, Ricoh Theta, Resolve 360

Achievement: Successful project launch to life, Visit here!.
Paper accepted in Proceedings of the 16th International Conference on Web Information Systems and Technologies (WEBIST 2020).

Open Source Code

Conference Presentation

AI Visualisation Tool for Covid-19 First Diagnosis

Overview: Members from GCU's Data Science Research Group, led by Professor Gordon Morison, took part in the Corona Hack. This is a week-long competition, organised by artificial-intelligence consultancy Mindstream-AI, where multi-disciplinary teams from all over the UK use their data skills to devise innovative techniques to combat COVID-19 using publicly available data. The entire event took place online using Slack and Microsoft Teams.

Challenge: The GCU team, which Maria was part, developed a solution to aid first diagnosis of COVID-19 in hospitals based on chest X-rays and clinical annotations. During the course of the week, the team members used their data science and software-development skills to create a prototype visualisation tool, which uses X-ray images and other patient data to show doctors the level of COVID-19 severity in a patient.

Scope: Five Days. From research to conception, prototyping.

Tools: Streamlit, Python, AI

Achievement: 2nd place Hackathon Competition.

2D/3D Visual Analytic Tool for Structural Condition Monitoring and Analysis

Overview: Large volume of complex data are collected which require visualisation and interaction techniques to enable domain experts to extract useful information. Artificial Intelligence (AI) technologies enable data science researches to develop AI algorithms to automatically process and analyse this data. However, these algorithms often output complex data that requires AI experts to interpret. This creates a gap between the AI researches and the domain experts, who are not familiar with AI but could benefit from its use.

Challenge: Development of a 2D/3D visualisation tool to support structural examiners when conducting a structural condition assessment. This tool allows to allocate structural defects, identify its type, and evaluate the severity level in a photorealistic 3D environment. A case study focused on structural examination of tunnels is presented to demonstrate the applicability of the proposed system.

Scope: From concept to research, design, prototype, and user testing

Tools: C++, Unreal Engine, Blender

Image Acquisition System for Structural Condition Monitoring

Overview: Railway tunnel inspections have been mainly undertaken in the traditional manner, which does not provide very reliable data and it is time consuming. Therefore, alternative methods need to be developed to have more quality in data collected from the tunnel inspections.

Challenge: Design and manufacturing an optical-thermal camera rig for railway tunnel inspection. This system is therefore to supplement the visual and manual procedure, being more effective and contributing to the maintenance of the railway sector.

Scope: Bachelor Thesis in twelve weeks. From research to conception, design, prototype, and user testing.

Tools: AutoCAD, SolidWorks, Keyshot

Achievement: This work is part of my team honours project and was marked as a distinction dissertation. The work was part of a paper called “An imaging system for visual inspection and structural condition monitoring of railway tunnels” written by my supervisors Dr. Mark Jenkins and Prof. Gordon Morison.

Visualising Water Environment in Scotland

Overview: Scotland's Centre of Expertise for Waters (CREW) project aims is to raise awareness to general audience and experts of the water environment around populated areas and its pharmaceutical concentration in water.

Challenge: Ideate Data Visualisation solutions to communicate information on pharmaceutical risk to different audiences from domain users to non-experts. This work was part of the first stage of Scotland CREW-Reduction project, where just visualisation recommendations were required.

Scope: Ideation, research, visualisation

Tools: Sketching, Vis Techniques, GIS

Web-based 2D/3D Visualisation

Overview: Many open-source tools like JavaScript library D3, Tableau and R exist to create interactive visualisation platform. Furthermore, there are other tools that allow rendering and interacting 3D graphics like Potree.

Challenge: Exploration of open-source for 2D and 3D interactive visualisation. Potree allowed discovering its potential tools to measure, annotate and extract the elevation profile from 3D reconstructed model. R and D3 allowed exploring common visualisation techniques such as parallel coordinate and boxplot graphs.

Scope: Ideation, research, visualisation

Tools: D3.js, Potree, ggplot2, RStudio