Students innovate to combat waste, dementia and allergies

Malaysian students create award-winning AI solutions addressing food waste, dementia support, and child allergy safety, showcasing the nation’s growing role in impactful, socially driven technology. Read full story

featured-image

THE recent Intel AI Global Impact Festival 2024 saw Malaysian students taking centre stage in the realm of artificial intelligence (AI), with projects that address critical challenges such as memory support, waste management and child allergy safety. The event, now in its fourth year, brought together young minds from 25 countries, all eager to showcase AI solutions aligned with the United Nations’ Sustainable Development Goals. Judged on creativity, social impact and responsible AI practices, these projects underscored the power of youth-driven innovation to tackle pressing global issues.

The event, which emphasised the theme “Bringing AI Everywhere For Everyone”, promotes AI that is inclusive, equitable and accessible. Among the 13-17-year-old category winners were Ong Chong Yao and Mok Zi Xuan, students from Han Chiang High School in Penang, who were celebrated for their AI-powered cabinet, the AI Storage Sentinel, designed to assist dementia patients and reduce food waste. Meanwhile, in the AI for Accessibility Award category, Iris Yan, a software engineering graduate, was awarded for her Pals (Pediatric Allergy Life Saver) project, an AI-driven tool aimed at improving the management of children’s allergies.



The recognition of Ong, Mok, and Yan at the Intel AI Global Impact Festival speaks to Malaysia’s expanding presence in AI and technology innovation. These young minds not only demonstrated their technical acumen but also their commitment to creating solutions that enhance the quality of life for communities in Malaysia and beyond. The AI Storage Sentinel and Pals illustrate how AI can address real-world challenges in ways that are responsible, inclusive and accessible.

Revolutionising everyday organisation Year 11 students Chong and Mok created the AI Storage Sentinel, an intelligent cabinet system that uses AI to keep track of household items. The duo developed their project with a clear mission: to assist individuals, particularly those with dementia, by reducing the anxiety and frustration that comes from misplaced items. Mok (left) and Ong developed the AI Storage Sentinel, an AI-powered cabinet designed to assist dementia patients and combat food waste.

— ONG CHONG YAO “People often feel stressed and confused when they can’t find an item quickly, especially those with memory challenges,” said Ong in an email interview. The AI Storage Sentinel uses a camera and machine learning algorithms to monitor items placed within the cabinet, identifying and tracking each item through real-time object recognition. “We used a Logitech camera with a custom-trained Yolov8x object detection model (Yolov8x is the latest iteration in the Yolo series of real-time object detectors) to identify objects in the frame,” explained Ong.

“The system also uses a mathematical function to track whether an item is being placed into or removed from the cabinet, with all movements recorded in real-time on a Firebase database.” This practical design allows users to easily monitor their belongings through a smartphone app, enabling them to check on the contents of their cabinets remotely. In addition to helping dementia patients, the AI Storage Sentinel is also an eco-friendly tool aimed at minimising food waste.

“With this system, people can be more mindful of what they already have, reducing unnecessary purchases that often lead to expired food,” Mok shared. “Our plan is to implement a feature to track the expiry dates of items, allowing users to use products before they spoil.” This aspect of the project, they shared, aligns with the UN’s Sustainable Development Goal 12, which emphasises responsible consumption and production.

Creating the AI Storage Sentinel posed several challenges. Ong and Mok faced tight time constraints as they developed their prototype during a crucial academic year, balancing Cambridge International General Certificate of Secondary Education coursework with their project. They also faced technical limitations, using just one powerful computer for development, which restricted real-time collaboration.

“Also, we had to self-learn all the skills and read relevant articles on our own too!” Mok said. Despite these hurdles, they succeeded in building a functional prototype that highlights the potential of AI to support memory-impaired individuals and promote sustainable living practices, earning themselves the grand prize of US$5,000 (RM22,350), an Intel-powered laptop, a certificate, and mentorship opportunities. What’s in the pipeline for this dynamic duo? “For now, we plan to decrease the overall cost of our project to make it more accessible to people of all incomes, while also making it more user-friendly for the elderly.

"Then, we want to include generative AI into our project to recommend recipes to users based on what they have in their storage,” Ong ambitiously summed up. Enhancing allergy management Yan’s project, the Pals app, was inspired by real-world observations and her love of hackathons, where she constantly searches for technology-driven solutions to everyday problems. Based in Selangor, the USM grad was particularly struck by the importance of allergy management after viewing an online video of a child suffering from a severe allergic reaction.

Yan designed the Pals app to utilise AI for ‘symmetric semantic search’, allowing users to easily find allergy information without needing precise medical terms. — IRIS YAN “I was scrolling on Instagram and came across a video of a toddler that was struggling to breathe, and it turned out to be an allergic reaction to bread! It occurred to me that even something as innocuous as bread can be a threat because just about anything can be an allergy – our bodies work in very mysterious ways. "It then got me thinking about the potential digitalisation of allergy management solutions and the level of allergy awareness in Malaysia.

"I spoke to a few friends who grew up with allergies to understand more about the problems they faced when they were kids and also attended a talk about allergy management held at my university to interview a certified pharmacist,” she shared about what prompted her to develop a tool to empower parents and caregivers in managing children’s allergies. Pals leverages AI technology to perform a “symmetric semantic search” for allergy symptoms, allowing users to find information without precise medical terminology. “Using Sentence Transformers (Sbert), Pals interprets the meaning of words and provides results relevant to the symptoms or concerns users input,” explained Yan.

“It essentially acts as a dedicated search engine for allergies, tailored for immediate assistance.” The system also includes an allergen detection feature that scans product labels to identify potential allergens, an essential function for children with multiple allergies. Using optical character recognition (OCR) technology, Pals extracts text from ingredient lists and detects allergens, assisting caregivers in determining if a product is safe for consumption.

“This feature is also accessible to visually impaired users who may find it challenging to read small ingredient labels,” Yan noted. Designing a reliable AI solution for health applications requires high standards of safety and accuracy, and Yan took meticulous steps to ensure this. She consulted a certified pharmacist to validate the allergy management features and incorporated a feedback loop into the system to monitor user interactions and continuously improve Pals based on real-world data.

Despite these safeguards, Yan is mindful of the limitations inherent to an AI-based health tool and clearly communicates that Pals should complement, not replace, professional medical advice. Yan’s vision for Pals extends beyond individual use. Recognising the need for consistent allergy management in childcare settings, she has integrated a classroom management feature into the system.

This allows caretakers to oversee groups of children with allergies, offering a searchable list of each child’s allergies. Additionally, Pals includes an AI-safe meal planner powered by the GPT-3.5 API, which suggests meals that cater to each child’s dietary restrictions, simplifying meal preparation for childcare staff.

Yan envisions future enhancements for Pals, such as integrating larger datasets to improve the accuracy of allergen detection and connecting parents with healthcare professionals directly through the platform. “AI will continue to transform healthcare, and I’m excited to explore more ways that it can make a positive impact on children’s lives,” she said. Meanwhile, as Intel continues to encourage responsible AI practices with events like the Global Impact Festival, Malaysia’s contributions to AI innovation are likely to grow, paving the way for future breakthroughs led by our innovative and industrious young local talents.

With students like Ong, Mok and Yan at the forefront, the nation is poised to make a lasting impact on the global AI landscape, furthering Malaysia’s reputation as a hub for creative and socially impactful tech solutions..