Australian Capital Territory

Year – 10

= First Place

PancVision- Pancreatic Tumor Segmentation Harnessing 3D UNet Architecture
Richard Ji
Burgmann Anglican School

This project involves the segmentation of tumors within the Pancreas. Using the 3D Unet Architecture in combination with another innovative algorithm, I aim to be able to accurately identify and isolate tumor regions within medical imaging data. This will enable more precise diagnosis and treatment planning for patients with pancreatic cancer.

= First Place

Breast Cancer Detection
Shree Jonnalagadda
Burgmann Anglican School

This project involves the segmentation of tumors within the Pancreas. Using the 3D Unet Architecture in combination with another innovative algorithm, I aim to be able to accurately identify and isolate tumor regions within medical imaging data. This will enable more precise diagnosis and treatment planning for patients with pancreatic cancer.

Second Place

Thumbtastic: A robotic replacement thumb
Michael Bartlett, Lani Otesile, Alex Larkings
Burgmann Anglican School

Our goal is to create a robotic third thumb that can be controlled using other body parts i.e.. big toes, head, other thumb. It will be attached to the hand by either using an elbow glove or elastic straps that surround the hand. Our idea is that the thumb will be controlled by muscle fibre that acts as tendons and muscles for the thumb. The alternative is using elastics connected to servos/motors to perform similarly. This will allow a range of motion and similar control to a real human thumb, so therefore can perform similarly complex tasks. The control of all the servos or muscle fibres will be controlled by a micro-bit that is wirelessly taking input from flexibly sensors connected to a second microbit. The controlling microbit and servos will be attached to the upper glove or a bracelet worn on the lower arm of the user.

Year 11 – 12

First Place

Laser Harp
Ethan Platt, Summer Saunders, Andrew Shobbrook, Matthew Thomas
Burgmann Anglican School

The project that will be submitted for this event is a harp, controlled via Arduino units, that is used via lasers to produce the strings. At the base of the harp, a line of photoresistors sit. This means that if the laser is interrupted, the light sensors can play a note with the attached buzzers.

The laser harp was created to make an easily affordable and adjustable harp that is able to perform just as effectively as a normally stringed harp. The harp will have at least one octave of notes, of which can be changed via the addition of a button into the circuit. The build itself will be modded to a 3D-printed plastic stand, allowing for the harp to have a level of protection against possible interference, as well as giving the harp an appealing aesthetic.

Second Place

AI Personal Workout Trainer
Muhammad Irsyad Yunus
UC Senior Secondary College Lake Ginninderra

AI-driven personal workout trainer designed to elevate home fitness routines through the integration of OpenCV and Python. The system leverages advanced computer vision and machine learning technologies to provide real-time exercise guidance, form correction, and personalized workout plans, enhancing the effectiveness of individual workout sessions.

The AI personal workout trainer utilizes OpenCV’s image processing capabilities to analyze users’ movements and postures in real-time. Through custom-built machine learning models, the system detects exercise-specific keypoints and assesses users’ form, ensuring correct execution to prevent injuries and maximize workout benefits. Visual feedback is provided on users’ screens, creating an interactive and immersive training experience.

In addition to form correction, the system offers personalized workout plans tailored to users’ fitness levels, goals, and preferences. The AI’s data-driven insights help create dynamic routines that adapt over time, optimizing progress and preventing plateaus. Users can also receive real-time performance metrics, encouraging accountability and motivation.

The applications of the AI personal workout trainer extend to various fitness levels and goals. From beginners seeking proper guidance to experienced athletes refining their techniques, the system provides comprehensive support. The integration of OpenCV and Python empowers users to engage in effective workouts without requiring specialized equipment or dedicated gym spaces.

Third Place

WINSTON
Samuel Sidebotham, Morgan Potter, Joshua Kisnorbo
Burgmann Anglican School

Our project is a robot dog that can navigate and map its environment. It is a combination of a number of different projects, each focused on a different part of the dog’s functionality. We’ve emphasised the physical build of the dog, using 3d printed parts, servo’s, gears, and timing belts to replicate dog leg movements. We’ve implemented an IK algorithm to enable the dog to control its movement more conveniently and provide it with commands to navigate its environment. We’re using computer stereo vision to allow Winston to map its environment, using a dense stereo matching algorithm to create a height map of the environment that can be displayed in real time. Lastly we’ll have a web interface that will allow you to see what Winston ‘see’s, and to give a system to control Winston.