Page 56 - FoodFocusThailand No.243 July 2026

Moreover, there is pressure shift freezing, which exploits conditions. For example, the freshness of salmon stored
the principle that the freezing point of water decreases under under fluctuating temperatures can be predicted using CNN-
high pressure. When pressure is suddenly released, freezing LSTM models, a type of deep-learning architecture capable of
occurs simultaneously throughout the food product, producing learning temperature variation patterns and microbial growth
fine ice crystals and significantly reducing texture damage. more effectively than traditional kinetic models. In the seafood
Magnetic Resonance Freezing (MRF) also represents an industry, electronic noses are commonly used to monitor
advancement in the industry. This technology uses magnetic ammonia and carbon dioxide levels in tilapia and white
waves to vibrate water molecules, preventing crystal formation shrimp, enabling freshness assessment within minutes. For
until the temperature drops below the normal freezing point. dairy products, machine-learning models have demonstrated
The magnetic field is then removed, allowing rapid and uniform high accuracy in analyzing proteolysis in mozzarella and
freezing throughout the product. Similarly, ultrasound- cheddar cheese, helping manufacturers identify the optimal
assisted freezing accelerates nucleation and controls ice consumption period before products reach consumers.
crystal size through cavitation effects. This approach is Moreover, studies on freeze-thaw cycles and temperature
particularly beneficial for delicate products such as fruits and abuse have shown that partial thawing followed by refreezing
soft-textured foods, helping prevent excessive hardness after can cause ice accumulation on packaging surfaces and lead
freezing. to undesirable texture changes in products such as mashed
potatoes and ready-to-eat pasta. In these cases, combining
Precision Shelf-life Analytics: IoT sensors with AI can provide immediate alerts when
Enhancing Shelf-life Prediction with AI products have been exposed to such conditions, preventing
Once food has been effectively frozen, accurately assessing the distribution of goods with compromised quality.
its shelf life under potentially fluctuating cold-chain conditions
becomes critical. Precision shelf-life analytics represents Smart Packaging and the Future of Intelligent
a shift from static expiration dates toward real-time shelf-life Food Supply Chains
assessment based on actual storage conditions. These technologies are no longer confined to research
Artificial Intelligence (AI), particularly Machine Learning laboratories. They are increasingly being commercialized
(ML) and Deep Learning (DL), has become central to through smart packaging systems that incorporate Time-
processing complex data generated from non-destructive Temperature Indicators (TTIs) and pH-sensitive labels,
testing techniques such as hyperspectral imaging, Raman which change color when chemical changes begin to
spectroscopy, and Electronic Nose systems. These sensors occur in food products. Such information can be connected
detect biochemical changes, including lipid oxidation and to smartphone applications or supermarket inventory
increases in volatile compounds associated with spoilage. management systems, enabling dynamic pricing based
The resulting data are integrated with mathematical models on the actual remaining shelf life. This approach helps
such as the Arrhenius equation and quality indices to estimate reduce food waste by allowing products that remain safe
the remaining shelf life more accurately than conventional and of acceptable quality to be sold before reaching their
approaches. conventional expiration dates. Furthermore, integrating
AI with cloud computing and blockchain-based tracking
From Research to Industrial Applications systems enhances traceability across the supply chain.
Numerous studies have focused on evaluating the quality of Manufacturers can monitor the temperature of shipping
frozen foods through Accelerated Shelf-Life Testing (ASLT), containers worldwide and adjust distribution plans when
in which products are stored under elevated or fluctuating products begin to deteriorate more rapidly than expected.
temperatures to accelerate deterioration. The resulting data Today, innovations such as HyFloFreeze™, which utilizes
are then used to predict performance under normal storage industrial-scale Hydrofluidisation technology, have already
been implemented in Europe. This technology reduces
energy consumption while preserving the quality of frozen
fruits and vegetables more effectively than conventional air-
blast freezing systems.
The integration of advanced freezing technologies
that preserve cellular structures with AI-driven precision
shelf-life analytics and smart sensors is becoming a key
direction for the future of ready-to-eat food production.
These innovations not only improve product quality,
safety, and consumer experience but also contribute
significantly to building sustainable food systems, reducing
food waste, and enhancing the economic efficiency of the
Thai food industry.

More Information Service Info C009

56 FOOD FOCUS THAILAND JUL 2026

23/6/2569 BE 18:31
52-56_��������� 1.indd 56
52-56_��������� 1.indd 56 23/6/2569 BE 18:31