Ai-powered hybrid modeling platform for optimizing ice cream texture and flavor

Overview

SENSE ICE is an innovative AI-powered modeling platform designed to enhance the texture, stability, and flavor of ice cream. This solution integrates mechanistic simulations with advanced machine learning to predict the effects of formulation and processing on ice cream's sensory qualities. By combining physics-based and data-driven models, SENSE ICE offers accurate, explainable, and scalable predictions that streamline product development, reduce trial-and-error, and support new product innovations such as clean-label and plant-based ice creams.

Technical specifications

Key features:

• Hybrid modeling approach: Utilizes mechanistic models for simulating physical processes like crystallization, air incorporation, and thermal gradients.
• Machine learning integration: Empirically predicts sensory outcomes by capturing nonlinear interactions, enhancing the predictive accuracy of sensory attributes.
• Forward prediction and inverse design: Provides both predictive insights and design capabilities for novel formulations and processing conditions.
• Pilot-scale validation: Ensures scientific robustness through pilot-scale experiments and sensory panels, refining model accuracy and reliability.
• Adaptive and scalable: Continuously improves with new data and is applicable across different production lines and product categories.

Technology readiness level

The SENSE ICE platform is currently at Technology Readiness Level 3, indicating that the solution has been validated in a laboratory environment with a plan for future pilot-scale testing and validation through sensory panels.