Laidlaw Research Project Brief
In this blog, I briefly talk about my upcoming Summer research project which is in collaboration with the Laidlaw Foundation at Trinity College Dublin!
Welcome back to my blogs! Sorry that it has been a while (had to get exams out of the way)! Nonetheless, I'm very excited to get back into writing these blogs again! If you didn't know, I was recently awarded a scholarship with the Laidlaw Programme at Trinity College Dublin. As a part of the application, I had to come up with a research proposal that was to be presented to a panel of judges from Trinity's Career Services. To be honest, I'm still quite shocked that I managed to get it (which is even more evident in this photo that I'm in with my fellow scholars).
With the programme, not only do I get to research over the Summer, but I will also be partaking in many Leadership/mentoring sessions which will help me become a better student and confident graduate after the programme is completed!
But enough of that, let me tell you about my research project!
The official title of my research project is, "Automated Procedure for the Microbial Analysis of Contact Plates." In many hospitals (more specifically, microbiology labs), visual inspections remain the most common form of hygiene auditing. In particular, bacteria on surfaces are audited by taking samples using contact plates, which are similar to Petri dishes. These plates are developed over 24 hours, where the colony-forming units (CFUs) formed are then counted manually.
This traditional method brings about a few problems. Firstly, it is time-consuming as microbiologists can only audit one plate at a time. Secondly, it is prone to human error as you are manually counting the CFUs by eye. Finally, there is a risk of infection/illness due to prolonged exposure to these plates.
My research project aims to tackle all three of these problems by creating a system that will automate the reading of these contact plates. The end goal by the end of Summer 1 and 2 is to have a prototype that can:
Take high-resolution images of the surfaces of 5-10 contact plates.
Calculate the number of CFUs using artificial intelligence and machine learning.
Take the incoming data, and create associated statistics and data visualisations (e.g. graphs)
By the end of Summer 2020 (Summer 1), I hope to have at least a trained model that can count CFUs from existing images with ~95% accuracy and at a sufficient speed. Afterwards, I will be looking into creating the prototype rig for the system and reaching out to various microbiology labs to test the technology in the field.
My supervisor for this project is Dr Conor McGinn, who has been supporting my ambitions for almost two years now. He is known for his research regarding "social robotics," in particular his "Stevie" robot which was featured on the front cover of the "TIME" magazine.
I'll try my best to blog about my research journey, as well as make a few video shorts about my work on my engineering Instagram! I'm looking forward to getting started and I hope you do enjoy the content I'll be pumping out! Until then, you can stay up-to-date by following my Instagram as always, @joeceengineering!
I hope you enjoyed the read as always, and I'll see you in the next one!