Functional Data Meets Mixture DoE in a Food Production Case Study (2026-US-30MP-2974)

Food texture plays a critical role in consumer acceptance, requiring careful control of viscosity to achieve the desired sensory experience. At the same time, food and beverage manufacturers must ensure products can be processed efficiently to minimize waste while maintaining quality and affordability. Achieving this balance becomes especially challenging when formulation changes are required.

This presentation describes a case study in which a necessary ingredient change highlighted significant viscosity‑related processing challenges. To address this issue, a mixture design of experiments (DOE) was conducted using key food ingredient factors. Viscosity was measured across the process using a rheometer, which mimics the full scale process. This generated functional data captured the full behavior of the system rather than relying on single‑point measurements.

The resulting functional data were analyzed using the Functional Data Explorer and Functional DOE platforms in JMP to identify ingredient interactions and processing sensitivities. Insights from this analysis led to formulation changes that successfully resolved the processing challenges. Finally, custom JSL scripting and advanced visualization in Graph Builder were used to validate these improvements through analysis of historian sensor data at the production facility.