Dr Eyad Elyan

Tell us something people might not know about you.

I joined Robert Gordon University in 2009 after completing my PhD in computer vision from the University of Bradford.

What do you enjoy most about your job?

I enjoy the challenges and every day you have a new challenge. What I really love most about it is that we have the opportunity to put into practice what we develop in our labs to solve real-world business problems. Can we predict if equipment would fail -in advance-? How can we detect fake news or a doctored video or image? What makes a good athlete? Is it possible to process complex engineering drawings automatically? How can we transform large volumes of utilized data into knowledge? These are the types of questions, which I face, and it certainly makes my job exciting, challenging and rewarding.

Why did you apply to become a Reader?

I enjoy research and the Readership is a research-intensive role. So this was an opportunity for me to expand on my research activities.

Tell us about your research

My research is focused on machine learning and vision applications. The overall aim is to analyse and model volumes of complex historical data to inform business practices and future decision-making processes. As humans, we have some form of cognitive bias, we are shaped by our knowledge and experience. However, using advanced mathematical models; we can provide more objective and conclusive results based on historical data. 

The work is mostly focused on transforming structured and unstructured content, like images and video, into actionable knowledge by building and evaluating advanced intelligent models.

For example, models can be created to detect components and symbols within complex diagrams and drawings to reduce the workload of engineers. A diagram that could take hours for someone to analyse and process its component parts could take minutes through the application of machine learning, and that’s without human intervention. Take into account the continued use of diagrams and you have a real cost saving for an organisation, as well as an increase in the accuracy of the output.

Models can also be built for procedural tasks where mistakes could cost millions of pounds or put the wellbeing of individuals at risk. These tasks usually involve the use of manuals. Take, for example, an offshore diver tasked with opening three valves. When he goes down, he learns about the helmet and such tasks using a paper manual and practice how to open the valves i.e. – half turn one way and counter clockwise the other, etc…. A very simple task, but closing the wrong one could cost you your life. Using augmented reality, a tablet or other device could detect the valves and show the diver how to rotate them with the information overlaid – it’s all visual and easy to understand for anyone. We ran the experiment with 60 of our students to open and close three valves in a certain way. Half used a manual and the other half used our augmented reality. The difference was massive. The people using the tablet were able to complete the task faster and more accurately since the angles to rotate through where overlaid.

There are so many tasks that could be simplified and be made safer with techniques like this. Industries are aware of the benefits and want to utilise the vast amounts of historical data available to them. We’ve generated a lot of income from funding bodies and by partnering with industries to do so. 

What are you most proud of?

I am very proud of the team I’ve been working with at the school and proud to be a member of the RGU family. I’m working closely with several industrial partners to solve real-world problems and over the past few years, I’ve managed to generate substantial income. This enabled me to recruit very talented people as research assistants, research fellows, knowledge transfer partnership associates and PhD students. Certainly, I am most proud of being part of this team.

Is there anything you’d like to add?

I feel very fortunate to be part of the RGU family. Certainly, I received a lot of support from my school and from the university. It's no secret that RGU is investing in research. For example we were the first to acquire the DGX-1 machine, which is a super computer allowing us to process millions of images in a few hours. This kind of environment will help and advance the research agenda in our institute.

Reader is a Grade 9 role which looks to promote those who have achieved a significant level of activity in Research. If you are interested in applying for a Reader position, take a look at the Annual Appointment Process document on RGyoU.