Developing a modelling culture
In addition to enhancing the classic nutritional models, the team is exploring the enormous potential emerging to develop machine learning models which can predict, forecast and optimize animal and economic responses. The opportunity is compelling: to create decision-making tools that can serve a diverse range of customers with different needs and digitalisation levels.
The team looks to develop a modelling culture throughout the Trouw Nutrition organisation that enables employees to recognise the value and benefits models bring to customers, as well as their limitations. We also focus on understanding the needs of our customers across a continuum of digitalisation (minimal to highly automated) to determine the appropriate modelling method.
Most important is our team’s active participation in designing, creating and validating algorithms to more accurately assess the nutritional value of ingredients and predict the response of animals across species based on their nutritional, environmental and health status. These algorithms will form the engine of digital tools that we intend to use as a means to an end, not the end itself.