Rapid acoustic-optical texture classification for jelly and structured foods

Overview

Valor Infrastructure Holdings is developing a rapid acoustic-optical texture classification method for jelly and structured foods containing particulates. The approach applies controlled low-amplitude acoustic or mechanical excitation to a prepared sample or finished product and records the resulting optical/video response, including deformation, damping, recovery, and vibration behavior. These response patterns are converted into numerical features and classified against approved quality ranges, providing an at-line or near-line pass/fail or graded texture assessment in approximately 5 minutes.

The method addresses a practical production bottleneck: full laboratory texture profiling typically requires sample separation, conditioning, specialized equipment, and long turnaround times. By capturing practical texture behavior directly from the product or a minimally prepared sample, this technique is designed to reduce reliance on slow laboratory analysis and enable faster quality decisions in food manufacturing environments.

Technical specifications

How it works:

• Controlled low-amplitude acoustic or mechanical excitation is applied to a jelly or structured food sample
• Optical/video capture records deformation, damping, recovery, vibration, and free-water-related response features
• Response patterns are converted into numerical features and classified against approved quality ranges
• The initial objective is classification of whether samples fall within or outside the intended quality range
• Acoustic-optical features can be correlated with existing reference texture measurements including gel strength, elasticity, adhesiveness, cohesiveness, resilience, and free water

Key benefits:

• Approximately 5-minute decision target, versus lengthy laboratory turnaround
• Designed for at-line or near-line deployment with simplified sample handling
• Does not initially need to reproduce every texture profile parameter, focusing on practical pass/fail classification

Technology readiness level

The technology is in an early development stage. The proposed first phase would define representative acceptable and out-of-range jelly samples and collect matching reference texture analyzer data. Subsequent evaluation would compare low-amplitude acoustic or mechanical excitation and optical response features against existing quality classifications and texture measurements to determine whether a fast predictive model can separate acceptable from unacceptable samples. If successful, the next phase would adapt the method for at-line or near-line manufacturing use with simplified sample handling and a 5-minute decision target. The technology is currently at the concept validation and feasibility stage, with laboratory validation and pilot deployment as the next milestones.