Nanyang Technological University (NTU) in Singapore has unveiled a groundbreaking artificial intelligence (AI) model designed to predict bacterial growth in food products, enabling precise tracking of freshness levels throughout the supply chain. Developed by researchers at the university's Future Ready Food Safety Hub (FRESH@NTU), this innovation promises to significantly cut food waste while enhancing food security in a nation that imports over 90 percent of its nutritional needs.

The model addresses a critical challenge in Singapore's food ecosystem, where inefficiencies in storage and distribution lead to substantial losses. By providing real-time insights into how pathogens like Salmonella proliferate under varying conditions, supermarkets and wholesalers can optimize shelf lives, adjust storage parameters, and manage inventory more effectively. This proactive approach shifts food safety from reactive measures—such as halting operations after contamination incidents—to predictive prevention, saving time, resources, and ultimately, food.

🔬 How the NTU AI Model Functions Step-by-Step

The AI model operates through a sophisticated machine learning framework trained on empirical data collected under simulated real-world conditions. Researchers mimic supply chain environments by exposing food samples to controlled variables like temperature, humidity, and time, while monitoring bacterial growth dynamics.

Step 1: Data Acquisition – Sensors track microbial proliferation on specific foods, such as pork, from slaughterhouse to retail shelf. For instance, moisture and temperature fluctuations are logged continuously.

Step 2: Model Training – This dataset feeds into machine learning algorithms, enabling the system to forecast contamination risks and freshness degradation.

Step 3: Predictive Outputs – The model outputs include estimated pathogen levels, recommended shelf lives, optimal storage tweaks (e.g., raising frozen storage from -20°C to -16°C without compromising quality), and stock rotation advice.

Step 4: Integration – Designed for seamless deployment in cold chains, it complements existing guidelines, potentially reducing energy costs by fine-tuning refrigeration.

In tests, the model has demonstrated potential to extend product usability, directly tackling waste at its source.

Meet the Minds: Professor William Chen and Dr. Youssef Ezzaky

Leading the project is Professor William Chen, Michael Fam Endowed Professor and Director of NTU's Food Science and Technology Programme, as well as FRESH@NTU. A pioneer in food safety and novel foods, Prof Chen's career spans precision fermentation, zero-waste processing, and risk assessment. His vision emphasizes proactive technologies to align production with demand, reducing inefficiencies.

Dr. Youssef Ezzaky, Research Fellow at FRESH@NTU, brings expertise in microbial biotechnology and predictive risk modeling. Holding a PhD, he has contributed to studies on Salmonella inactivation in alternative proteins, blending mechanistic and machine learning approaches. Together, they exemplify NTU's commitment to interdisciplinary higher education research, training the next generation in AI-food tech fusion.

Singapore's Food Waste Crisis: The Numbers

Singapore generated 784,000 tonnes of food waste in 2024, comprising 12 percent of total waste, with recycling rates climbing modestly to 18 percent from 13 percent in 2014. Globally, one-third of food production is wasted, contributing up to 10 percent of greenhouse gases. In a land-scarce nation like Singapore, this translates to economic losses and heightened vulnerability amid supply disruptions.

NTU's model could reduce retail waste by enabling precise shelf-life predictions, potentially saving millions in discarded goods annually. For context, households spend around S$1,800 monthly on food, amplifying the stakes.

NEA's latest waste statistics underscore the urgency, positioning university-led AI as a vital tool.

Boosting Food Security: Aligning with '30 by 30'

Singapore's '30 by 30' initiative aims for 30 percent local nutritional production by 2030, countering 90 percent import reliance. NTU's contributions, via FRESH@NTU—a tripartite hub with A*STAR and SFA—extend to novel foods like cell-based meat and precision fermentation.

The AI model supports stockpiling and cold chain resilience, ensuring buffer stocks remain viable longer. Prof Chen notes it enables energy savings— a one-degree shift slashes bills significantly—freeing resources for security enhancements.

Industry Collaborations on the Horizon

FRESH@NTU is negotiating trials with supermarkets like Sheng Siong, set for late 2026. This aligns with industry priorities for safety amid rising consumer demands. Cold storage firms could adopt it for optimized operations, reducing poisoning risks and waste.

Broader partnerships, including WHO for AI-driven assessments, position NTU as a global leader. CNA reports highlight Sheng Siong's enthusiasm.

FRESH@NTU: NTU's Innovation Engine

Launched under Singapore Food Story R&D, FRESH@NTU pioneers safety for 'future foods' like insects and cultured proteins. AWS collaborations advance cloud-based predictive analytics, while education programs train experts in whole genome sequencing (WGS) and AI.

• Risk assessment for novel ingredients
• Consumer empowerment via communications
• Industry ecosystem building

Technological Edge and Scalability

Building on NTU's legacy—like the 2020 AI 'electronic nose' for meat—the new model integrates IoT sensors for supply chain monitoring. Dr Ezzaky's Salmonella modeling complements it, achieving high predictive fidelity via hybrid mechanistic-ML methods.

Scalability to seafood, veggies, and beyond is planned, with cloud deployment minimizing hardware needs.

Careers in AI-Food Tech at Singapore Universities

NTU's breakthroughs spotlight opportunities in food science, AI, and biotech. Roles span research fellows, data scientists, and professors, with FRESH@NTU hiring amid '30 by 30' push. Similar at NUS and SUTD, fostering interdisciplinary talent.

Photo by Catgirlmutant on Unsplash

Stakeholder Views and Future Outlook

Prof Chen envisions market-ready tools by 2027, expanding to farms for demand-aligned production. Industry praises the shift to proactivity, while policymakers see it bolstering resilience. Challenges like data standardization remain, but NTU's ecosystem positions Singapore at the forefront.

In summary, this AI model not only curtails waste but elevates higher education's role in national priorities, inspiring global adoption.