ViscNet: Vision-Based In-line Viscometry for Fluid Mixing Process

2025-11-30

AI Summary

Conventional viscosity measurement relies on invasive sensors and idealized laboratory conditions, failing to meet the requirements for online, non-contact monitoring in industrial processes and autonomous experimentation. This paper proposes a computer vision–based, non-contact, online viscosity sensing method that exploits optical refraction-induced distortions of background patterns caused by free-surface deformation during fluid mixing. Leveraging a multi-pattern encoding strategy and an uncertainty quantification mechanism, the approach enables robust viscosity inference under complex illumination and realistic operating conditions. The method integrates deep learning–based regression and classification, physics-informed optical distortion modeling, and confidence estimation. Experimental results demonstrate an average absolute error of 0.113 on the logarithmic viscosity scale and a classification accuracy of 81%, significantly improving discrimination among adjacent viscosity classes and prediction robustness. The framework exhibits strong potential for automated adaptation to diverse fluid systems and operational environments.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multi-instance/Multi-view Learning

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Virtualization and resource management in Web systems and infrastructures
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Viscosity measurement is essential for process monitoring and autonomous laboratory operation, yet conventional viscometers remain invasive and require controlled laboratory environments that differ substantially from real process conditions. We present a computer-vision-based viscometer that infers viscosity by exploiting how a fixed background pattern becomes optically distorted as light refracts through the mixing-driven, continuously deforming free surface. Under diverse lighting conditions, the system achieves a mean absolute error of 0.113 in log m2 s^-1 units for regression and reaches up to 81% accuracy in viscosity-class prediction. Although performance declines for classes with closely clustered viscosity values, a multi-pattern strategy improves robustness by providing enriched visual cues. To ensure sensor reliability, we incorporate uncertainty quantification, enabling viscosity predictions with confidence estimates. This stand-off viscometer offers a practical, automation-ready alternative to existing viscometry methods.

Problem

• Develops a non-invasive vision-based viscometer for fluid mixing
• Measures viscosity using optical distortion from a deforming free surface
• Provides uncertainty quantification for reliable process monitoring automation

Innovation

• Vision-based viscosity measurement via optical distortion
• Multi-pattern strategy enhances robustness with visual cues
• Uncertainty quantification ensures reliable viscosity predictions