ISSUE FOCUS FEED & ADDITIVE MAGAZINE April 2024 45 PROBLEMS ASSOCIATED WITH NUMEROUS FORMULATIONS AND SENSOR CALIBRATION The major issue associated with multiple formulations, as required by Aquafeed manufacturers, is to ensure correct calibration of sensors for each formulation, or family of formulations. It is impossible to manage these calibrations if using a sensor that varies its output due to material or ambient temperature variations, wear and tear, or other installation issues such as an optical sensor becoming contaminated by dust or being affected by vibration. As the number of sensors used across the process increases it becomes even more important that each one works accurately and that all sensors used have a repeatable and reliable output and have identical measurement characteristics. A significant advantage of the Hydronix sensors is that each sensor will output an identical default factory calibrated value. This may be used as a direct linear input into a standard PLC feedback control loop and does not require any material calibration. The requirement for an absolute moisture value, requiring calibration, is redundant. Advanced independent system integrators are now able to use sophisticated AI-based algorithms linked to Hydronix sensors to continuously monitor and fine-tune the calibrations in real-time. Using stable sensors that all have identical measurement characteristics allows users to reliably record and re-use calibrations. One further indirect advantage of this is that any sensor may be removed and replaced without any need for re-calibration. The Hydronix sensor’s default linear measurement can output a value between 0 and 100 units in increments of 0.1. This output is known as the Unscaled Value and is identical between sensors. A user can simply set a target Unscaled Value for each formulation. If production switches between formulations, it is simple to change to the appropriate Unscaled Value target. Using sensors with identical output characteristics makes it possible for precise control loop control, enabling more and more users to adopt AI generated control loops for the best possible performance. CONCLUSION With the Hydronix measurement, there is confidence that the measurements are precise, accurate, and repeatable, allowing their use to improve the AI model and constantly improve efficiencies. This allows the feed producers to achieve consistency in their process, whilst working with many formulations, reducing variations in the individual process stages, and allowing them to target a higher final product moisture, whilst concurrently improving quality. Quality testing by one of Hydronix’s automation partners demonstrated a 40% reduction in Standard Deviation, achieving +/-0.4% moisture. This allowed their customer to increase their final moisture target to 10.6% to ensure product is always below 11%. Overall, Hydronix moisture sensors offer accurate, precise, and repeatable measurement, giving an improved and consistent product, a reduction in spoiled and wasted material, better plant efficiencies and reduced process downtime. The team here at Hydronix would welcome any further questions. References 1. https://www.researchgate.net/figure/Average-percentage-of-moisture-content-in-different-fish-feed-ingredients_fig1_327137357 About Neal Cass After gaining his degree in Electronic Engineering from the University of Southampton in England, Cass spent 10 years developing and commissioning control systems for a major international food process system manufacturer. In 2007 he started working for Hydronix as a Customer Service and Software Development Engineer before becoming Sales Manager in 2011.
RkJQdWJsaXNoZXIy MTUxNjkxNQ==