Call for Partners | UHT-PREDICT: Predictive modelling of Stability of UHT Drinks

Physicochemical stability is a critical quality parameter for UHT dairy and plant‑based beverages, where fat–protein emulsion performance strongly influences shelf-life, visual quality, and consumer acceptance. Currently, stability can only be verified through long real‑time storage tests lasting months to several years, making both product development and routine quality control slow, costly, and uncertain. Instability often emerges months after production, and batch‑to‑batch variations remain difficult to predict due to the complexity of the underlying mechanisms.
UHT‑PREDICT addresses this challenge by developing a fast, science‑based framework for anticipating physicochemical instability.

The project will combine accelerated testing, high‑throughput physicochemical measurements, and advanced data analysis to map the key drivers of creaming, aggregation, and phase separation in complex UHT systems.

This will be achieved by developing a hybrid predictive model that integrates data‑driven AI with mechanistic understanding. We will combine experimental datasets with AI and physicochemical models of complex emulsions and dispersions to provide accurate predictions that align with the underlying physicochemistry of emulsion stability. This hybrid framework will allow early identification of stability risks—within days rather than months—and expand applicability across diverse formulations and processes.

The goal is to create a robust, predictive model that supports formulation optimization, process adjustment, and proactive quality control. This will shorten development cycles, reduce cost, and increase production reliability across both dairy and plant‑based UHT products.

Recent advances in automation, rapid measurement technologies, and machine‑learning approaches now make it feasible to generate high-quality datasets on emulsion behaviour. UHT‑PREDICT leverages these capabilities to move beyond empirical, time‑consuming storage trials and toward fast, predictable stability management for UHT beverages.