Exceptional young scientists examine deepfakes, quantum computing and more

The judges at this year’s BT Young Scientist and Technology Exhibition face a huge challenge in finding the best among these spectacular student projects.

A BT Young Scientist and Technology Exhibition like no other is about to come to a close.

The first ever virtual edition of the event has been running this week, meaning that we could still start the year being wowed by the talent of Ireland’s next generation, with attendees from around the world offered access via an online portal.

Anticipation is mounting as to which project will be crowned the winner of the 2021 event and, with more than 500 to choose from, the judges have their work cut out for them.

Covid-19 featured as a prominent theme across many projects, as did sustainability and the climate crisis. Other teen scientists studied what they know best, examining the impact of push notifications and apps such as TikTok.

One student from Dublin’s Stratford College set about creating a higher-level programming language for quantum computers.

High-level programming languages are more accessible and require less knowledge of computer hardware. Transition-year student Ayush Tambde wanted to create one such compiled programming language for quantum computers, which could be read and written with ease and succinctness.

In his project video, Tambde not only demonstrated how he achieved this but also proved himself a stellar communicator in making quantum computing itself easy to understand. He said that coding quantum algorithms without a high-level programming language would be “like trying to build a house with nothing but a screwdriver”.

“That’s why I made Q. It’s a language that uses all these tasks,” he said.

Results presented by Tambde from an IBM simulator suggest that Q works, and he intends to open-source it so that anyone can try it out.

‘My laptop was starting up and it was slow and I decided to make an operating system to make it faster’
^{– PIOTR STELMASZEK}

Another student experimenting with code is Piotr Stelmaszek from St Kilian’s German School in Dublin. This junior-cycle student wrote an astounding 6,000 or so lines of code in Rust, C and Assembly for his project to design a better operating system.

“The reason I made it is my laptop was starting up and it was slow and I decided to make an operating system to make it faster,” he explained. “I think there should be more availability than just three big options (Windows, Linux, Mac). There must be something better.”

Stelmaszek was pleased with his results, producing an operating system of 16MB that boots in less than 75 milliseconds, compared to a Linux boot time of more than 400 milliseconds. “That’s like over four times faster, if I paid attention in maths class,” the young scientist cheerfully explained.

Bandon Grammar School student Greg Tarr set out to put powerful computing power to work detecting deepfakes.

“My project uses artificial intelligence to detect deepfake videos. This is a very active field of research where my project was able to speed the process up by a factor of 10 while retaining state-of-the-art accuracy,” he declared with great confidence.

To do this, Tarr studied the top five entries from a million-dollar contest to improve deepfake detection. He found even these winning projects could benefit from some improvement and set about rewriting the media loader, refactoring the code and training a face detector over hundreds of hours in order to make a more efficient and cost-effective system.

“Deepfakes have already had a major detrimental social impact, influencing elections, embarrassing people with the generation of fake explicit images and even conducting fraud,” said Tarr. “This means that large-scale automated deepfake detection can now be deployed at a tenth of the cost.”

Stepping away from computing and deeper into mathematics, Louth student Peidi Li conducted an investigation into regular and semiregular tessellating patterns for her Young Scientist project.

The fifth-year student from Our Lady’s College set out to investigate the existence of an irregular convex pentagon that has not yet been established as a regular tessellation, and to create her own tessellating designs.

In the simplest of terms, Li explained, tessellations use different shapes to tile a plane. There are many examples of this found in nature, such as the beehive’s hexagons and the patterns formed in the cracking of lava. These patterns can be used for aesthetic purposes in design and for structural purposes in engineering, to improve strength and stability.

“We are connected to tessellations more than we think, and a better understanding of tessellations will help you to improve your understanding of a two-dimensional and a three-dimensional space,” said Li.

Chiara Lisa Carra and Elysia O’ Leary with their project: ‘Smart Protective Sporting Gear’. Image: Chris Bellew/Fennell Photography

Another project finding everyday applications for advanced science came from Chiara Lisa Carra and Elysia O’Leary, transition-year students at Nord Anglia International School Dublin. This duo set out to create agile and light protective gear for athletes using materials science.

To do this, they impregnated fabrics with shear-thickening fluids as well as additives such as graphene in order to develop protective gear that could absorb up to 100pc of impact energy.

“The gear itself will be made out of fibre-reinforced polypropylene front and Dyneema backing,” said O’Leary. “From the treated fabric, we will make everyday clothing items used for sports, such as socks, and reinforce them with extra shear-thickening fluids. These reinforcements will provide more protection to bones and vulnerable areas as well as provide support for the spin, ribs and joints.”

“Frequent injuries are something we encountered as athletes,” explained Carra.

“We hope to promote the wearing of protective clothing in sports and/or in daily circumstances such as cycling to reduce the deaths and injuries caused by sporting incidents every year by providing an option that does not reduce performance,” added O’Leary.

These projects and hundreds more have been thoroughly evaluated over the past few days ahead of the winners’ announcements this afternoon (8 January), with many prizes to give away.

“Our virtual judging process is well underway, and each project is judged a minimum of three times by some of our 80 volunteer judges before final deliberations ahead of our awards ceremony,” said judge Prof Orla Feely.

“It’s a great experience to hear more from students about why they wanted to showcase their projects at this year’s virtual exhibition. A number of them mentioned how grateful they were that the exhibition took place this year, and that they found that working on their projects helped them through the periods of restrictions associated with the pandemic.”

The public can watch the 2021 BT Young Scientist awards ceremony via the event’s online portal and on Facebook Live today at 1pm.