Winners

Stem cell therapy holds tremendous promise for treating illnesses by regenerating body tissues, but struggle to be approved due to unsafe and inefficient delivery methods. To address this, the Soft MicroRobotic “nichE” (SMORE) system was developed, which employs soft microrobots that mimic the body’s stem cell reservoirs to deliver large amounts of stem cells while promoting their healthy survival and integration. The microrobots were 3D-bioprinted from liquid cell-laden inks composed of a biocompatible polymer, a light-reactive initiator, magnetic nanoparticles, stem cells, and niche components. The material and formulation process was found to be suitably porous, elastic, and biocompatible for supporting live cells and microrobots demonstrated injectability and responsivity in navigating through a spinal model. As such, the system exhibits potential for overcoming the primary obstacles to efficient cell delivery, offering a pathway towards highly effective stem cell therapies that can transform the treatment of countless diseases. My Projectboard.

Less

What if a simple voice recording could diagnose Parkinson’s disease? To explore this possibility, I created an AI-powered diagnostic tool that analyzes voice recordings, detects subtle vocal changes linked to Parkinson’s, and formulates a diagnosis. These patterns are typically too minuscule for the human ear to detect, but the AI model, built using a sequential pipeline of neural networks, achieved a 91.11% diagnostic accuracy. What makes this tool especially powerful is its explainability. Beyond offering a binary diagnosis, the AI model generates a detailed, case-by-case breakdown of the steps and factors considered in each patient’s diagnosis. This depth of insight empowers precision medicine, allowing physicians to move beyond one-size-fits-all approaches and personalize care strategies based on each patient's distinct neurological fingerprint. This technology has the potential to revolutionize Parkinson’s diagnosis, offering a free, noninvasive, and accessible solution that works anywhere with a microphone and internet. My Projectboard.

Less

Probing Galaxy Disk Sizes Using the Largest Ever Catalogue of Ca II Absorbers Quasars are extremely bright objects powered by supermassive black holes. As their light travels across the universe, it passes through gas and dust in galaxies, which imprint distinctive absorption lines on the quasar’s spectra. Ca II-bearing gas stands out for tracing dusty, metal-rich, star-forming regions, providing valuable insights into galaxy evolution. However, Ca II absorbers are rare and challenging to detect, with only around 1,000 previously known in total. Traditional detection methods are inefficient, requiring significant manual labor. To address this, my project improves detection accuracy and precision by innovating an ensemble learning approach, where multiple machine learning models make better decisions together. This resulted in the compilation of the largest ever Ca II catalog, with over 1,600 systems. Since Ca II is located in galaxy disks, this expanded dataset also shows evidence that galaxy disks are growing in radius over cosmic time, revealing previously uncharted aspects of galaxy evolution. My Projectboard.

Less

VisionnAIre: une approche IA innovante et prometteuse pour un diagnostic oculaire précis et accessible La déficience visuelle touche 2,2 milliards d’individus mondialement, dont 1 milliard de cas pourraient être évités. Le manque d’accès aux diagnostics précoces et précis est un enjeu majeur, particulièrement dans les régions à faibles ressources, limité par les coûts élevés des équipements, la pénurie de spécialistes, et les méthodes lentes et imprécises. J'ai alors développé VisionnAIre: un système de diagnostic oculaire accessible et efficace qui repose sur la conception d'un prototype de rétinographe portable ultra-économique, le traitement de bases de données, l'entraînement de réseaux neuronaux convolutifs pour l'analyse d'images oculaires, et la création d'une application intuitive complémentaire. Actuellement, mon modèle IA reconnaît 8 pathologies majeures ainsi que des yeux sains avec >97% d’exactitude, surpassant l'exactitude des spécialistes humains. VisionnAIre propose un outil prometteur pour transformer l’accès aux soins oculaires, particulièrement dans les régions défavorisées. C'est un projet ouvrant de nouveaux horizons pour une médecine plus efficace, intelligente et équitable. Mon tableau de bord de projet

Less

Building a prototype of active response glasses to prevent photosensitive episodes such as seizures. The project uses an Arduino-based circuit and code to detect the signature of a strobe light and to activate a hacked LCD panel from 3D glasses, When activated the panel darkens which prevents the strobe light signals from reaching the brain thus preventing the photosensitive episode. My Projectboard.

Less

Wildfires and forest fires burn thousands of hectares of land each year. This produces large plumes of toxic gases that can contaminate homes downwind of the fires. Other factors influencing AQI (Air Quality Index) include organic compounds, furniture outgassing, household gases such as propane or smog, etc. We measured the outdoor and indoor AQI using a pair of embedded microcontrollers equipped with Wi-Fi, a dedicated short-range radio link, a solar charging system with a Li-ion battery, and BME688 gas resistance sensors and a SCD30 CO2 sensor. When the indoor or outdoor air quality becomes dangerous, our device send you a text message to alert the user to open or close their windows. In modern homes, this signal could be interconnected to their home air exchange system to prevent poor air quality from entering the home. This device will have an impact on millions worldwide, alerting us when our noses cannot. Our Projecboard.

Less

The overuse of chemical fungicides in agriculture has led to widespread off-target deposition, environmental contamination, and increased human exposure to carcinogenic compounds. This project presents an RNA interference (RNAi)-based formulation as a safer, biodegradable alternative for fungal disease control. RNAi works by silencing specific fungal genes at the mRNA level, blocking the production of proteins essential for growth and infection. Unlike traditional fungicides, this approach is target-specific, non-toxic, and does not accumulate in ecosystems. Its modular design allows rapid adaptation to different pathogens and agricultural contexts. The same gene-silencing platform is now being adapted for medical applications, including the treatment of antibiotic-resistant infections such as MRSA. This work advances a new class of RNA-based solutions that are both environmentally sustainable and medically relevant. My Projectboard.

Less

Toxoplasmosis is a common food-borne illness caused by the parasite Toxoplasma gondii, a parasite often associated with cat excrement. Through autophagy, the parasite can persist in the host's tissues for life (Pariyamon et al., 2024). Currently, no treatment options exist for the chronic phase of infection in the central nervous system. In immunocompromised individuals, reactivation during the chronic stage can be deadly. T. gondii Cathepsin L (TgCPL), an important protein in autophagy, has been implicated as a novel target for the treatment of chronic toxoplasmosis (Ortiz et al., 2025). However, the presence of the blood-brain barrier (BBB) makes drug development difficult. An integrated virtual screening framework combining Quantitative Structure-Activity Relationship (QSAR) modelling and bioinformatics tools was created to identify promising drug candidates. The framework prioritized compounds showing selective binding with TgCPL, BBB-permeability, adherence to Lipinski's-Rule-of-5, and no neurotoxicity. Compound library screening identified several compounds as promising candidates for in-vitro testing. My Projectboard.

Less

This project introduces a novel approach for continuous low-cost atmospheric monitoring in remote areas using micro super pressure (pico) balloons. Traditional High-Altitude Balloons (HABs), are expensive and often have limited spatial coverage. Therefore, it is challenging to obtain continuous, real-time data due to their limited flight time and high cost. In contrast, my PicoTrack solution uses inexpensive party balloons to provide a low cost and eco-friendly monitoring system for near-stratospheric altitudes which will complement HAB technology. PicoTrack has a modular electronic device with the capabilities to connect a variety of sensors which is carried by pico balloons. This design efficiently uses solar power allowing for long duration flights and continuous telemetry transmission. Several iterations of PicoTrack were tested, each utilizing various environmental sensors, allowing the possibility to study dynamic interactions between the surface and atmosphere. PicoTrack can be utilized and modified for different purposes to support environmental monitoring. My Projectboard.

Less

Earth’s waterways are a precious resource, and the need for large-scale monitoring of these environments has become critical. One approach relies on innovative biomimetic robots that exploit the evolutionary characteristics of native marine animals. This project involves the development of an underwater bionic robot inspired by the unique swimming kinematics of green sea turtles, for the autonomous monitoring of aquatic ecosystems. Locomotive capabilities were implemented on the device using a custom-designed multi-axis servo propulsion system, and optimized for efficiency using computational fluid dynamics simulations and empirical testing in a backyard pool. A dynamic depth control mechanism and autonomous swimming gait were achieved using a variety of electronic sensors and control loop algorithms, allowing for a predefined search routine. Ecological monitoring of coral health was demonstrated using a custom-trained AI-based imaging system with a detection accuracy of 96%, proving the viability of the robot in helping to protect sensitive marine ecosystems. My Projectboard.

Less

When lives are at stake, and every second counts, most search and rescue drones are too costly, cumbersome, and hard to deploy quickly. That's why I built TALON: a compact, affordable search and rescue drone that can lift off vertically like a quadcopter, then transition to fly fast and far like an airplane. After extensive research, I settled on a custom tiltrotor configuration that delivers the best of both worlds: increased speed and endurance, while also taking off from a patch no larger than a dinner plate. Using 3D printing, open-source electronics, and a DIY approach, I built TALON for a fraction of the cost of industry standards. Rigorous testing showed that TALON flies over 10× farther and 3× longer than similar drones—all under $1,500. TALON proves that high-performance tech doesn't need to break the bank: not just a drone, but a concept, and way forward. My Projectboard.

Less

Each year, over 69 million individuals suffer a traumatic brain injury (TBI), yet recovery trajectories remain unpredictable. We identified Brain-Derived Neurotrophic Factor (BDNF) as a key biomarker of neuroplastic capacity and conducted a meta-analysis of 60+ peer-reviewed studies to quantify its relationships with age, sex, BMI, depression, and acute exercise. Using regression modeling and a multi-factor confidence interval intersection, we developed the Neuronal Rehabilitation Index (NRI)—a novel predictive tool estimating individualized BDNF levels. NRI can be integrated into routine annual check-ups, enabling healthcare providers to assess a patient’s neuroplastic capacity and to develop more personalized rehabilitation programs for TBI patients. To increase accessibility and impact, we launched NeuroplasticityHub, a website featuring our NRI calculator, educational literature, and curated research resources to support recovery-oriented care. Our Projectboard.

Less

Many highway tanker truck rollovers are caused by liquid sloshing during abrupt maneuvers, posing serious safety risks to drivers and other road users and often resulting in fuel spills that lead to severe environmental pollution and costly cleanup efforts. To address this problem, I created a scaled-down model using a modified remote-control car and a custom trailer to investigate and test sloshing and rollover risks with different baffle designs by adopting 3D modelling and printing with AI-assisted analysis. My novel baffle design with Stacking Static Propellers in Opposite Directions (Smart Baffles) was developed to significantly reduce sloshing, improve tanker stability, and lower rollover risks when tested with water. It dissipates sloshing energy by creating turbulence and redirecting liquid flow into a controlled spiraling motion. This innovation paves the way for a safer, cleaner, and more sustainable fuel logistics industry, ultimately saving the lives of road users and protecting the environment. My Projectboard.

Less

Pathogen AI: Using AI to detect fungi in your food. Have you ever seen mold on bread and wondered if it is harmful, if you can still eat it or what to do if you did accidently eat it well my goal was to answer some of those questions using AI and eventually use my tool to help the healthcare system. I trained a google teachable machine with images to recognize and identify 5 common harmful fungi. In the end my program was identifying my selected fungi with nearly 100% accuracy. My project is just a small thing however eventually AI could be a great help in the medical field, agriculture and so much more. My Projectboard.

Less

Visual Drug Interaction Checker Polypharmacy, the concurrent use of multiple medications, is common among older adults and individuals with chronic conditions. However, patients taking five or more medications face a significant risk of adverse drug events, which rises further as the number of drugs increases. The complexity of drug interactions, especially in polypharmacy, can pose a major challenge to patient safety, yet no user-friendly tools exist to visually represent these drug interactions. This project addresses this critical gap by developing an interactive visual drug interaction checker tool which also provides additional information about the different interactions and its probable causes. This tool can be useful to both patients as well as healthcare professionals, improving patient care and medication safety by simplifying the complex relationships between various drugs. My Projectboard.

Less

The SmartSleeper: Optimize Your Sleeping Environment Sleep is one of the most vital aspects of human health. Because of this, I wanted to create a device that could help people achieve a more restorative sleep. While there are many products on the market that track sleep—such as Fitbit, Apple Watch, and Whoop Strap—they don’t actually help improve your sleep. I wanted to make something that didn’t already exist and so I decided to create a device that optimizes your room’s sleeping environment. After doing some research, I discovered that the most important environmental factors affecting sleep quality are temperature, humidity, air quality (including pet dander, dust, and pollen), blue light, ambient light, and ambient sound. Using scientific studies on sleep, I identified the ideal ranges for each of these metrics and built a device that measures each and provides actionable insights in a clear, easy-to-understand way. My Projecboard.

Less