More insights at the feed bunk
Ration formulation is currently based on only a limited number of measurements from the silage. To enable more precise feeding and better control of the feeding process, a NIR-sensor installed at the feed mixer may offer a solution. The key question is how reliable these measurements are for different feed mixtures and individual feed components. In this validation study, it was assessed whether the sensor provides meaningful and practical information for on-farm use, and where variability occurs. Also, it was assessed whether the ration was mixed accurately, according to the corresponding software. The results provide a clear picture of current performance and offer concrete guidance for further improvement of NIR-based feeding support in the future.
Broad research question
What do you need to be certain about when measuring in the mixing wagon
The study focuses on the accuracy of NIR measurements of dry matter, crude protein, crude fibre, crude fat and crude ash. Measurements were carried out in mixtures of grass silage, maize silage, by-products and a complete ration. This makes it clear in which situations the sensor provides usable information that can be directly applied in ration management and process control.
Approach
Comparing with laboratories provides a reference
For the validation, various feed samples were measured using the NIR sensor in the feed mixing wagon and compared with results from six laboratories. This made it possible to identify differences in component levels. At the same time, the homogeneity of the ration was assessed by analysing particle size distribution and feed value. In this way, both composition and homogeneity were evaluated.
Objective
Reliable measurement in practice is the standard
The use of NIR only adds value if measurements can be trusted across different mixtures and in daily operations. The objective is to determine whether the sensor is sufficiently accurate for assessing feed composition and quality and whether it can be used to monitor mixture homogeneity. This helps safeguard mixing quality and supports well-founded decisions.
Results and reflection
Currently usable for dry matter and protein further calibration needed
The sensor performs well for dry matter and crude protein in mixtures of grass silage, maize silage and by-products. For crude fibre, crude fat and crude ash, and for single-component mixtures, results are variable. It is therefore too early to rely on all absolute values. Further calibration will improve reliability.
Successful outcomes:
Reliable values for dry matter and crude protein in mixed rations containing grass silage, maize silage and by-products.
Practical control of mixing quality through particle size distribution and feed value analysis.
Lessons learned:
Values for crude fibre, crude fat and crude ash show variation.
Measurements in single-component mixtures are not yet sufficiently consistent.
Further sensor calibration is required before absolute values can be fully trusted.
For practice, this means applying the technology where it already works and continuing development where needed. Within NXTGEN Hightech Agrifood, we work with partners such as Trioliet to broaden applicability, improve calibrations and accelerate validation across the value chain. Step by step, NIR in the mixing wagon is evolving into a reliable link in the daily feeding process.