AI and Supercomputing Team Up to Stop Fish Feed From Shattering in Transit
· Source: bioengineer.org
In short
A research team at Islamic Azad University's Majlesi Branch has developed a hybrid framework combining laboratory experiments, CFD-DEM simulation, and an artificial neural network surrogate to map how aquaculture feed pellets are damaged during suction-type pneumatic conveying. Testing twelve commercial pellet types across three species, the study identifies flow-regime thresholds, a U-shaped pressure-drop optimum near 15–16 m/s, and damage rates reaching 8% for fragile pre-starter pellets. The resulting interpretable AI model predicts pressure drop, particle velocity, damage, and power consumption with R² above 0.96, yielding actionable design guidelines for aquaculture feed transport.
Read Original Article — Read the full article on the publisher’s site.
In the issue of week 40, 2026: Supply Chain AI Advances, but Data, Governance and Capacity Set the Pace
Tags: pneumatic conveying, aquaculture, CFD-DEM, neural network, feed pellets, simulation