Feed As A Service: Predicting Feed Orders At Dairy Farms

Twan van den Broek

Dairy farms need to order a lot of animal feed. Generally, farmers order when they check the silo – either by carefully looking at the slightly see-through storage tower or by tapping it to check how full or empty it sounds. They’re inventive techniques, but not very reliable, a fact that ForFarmers, an animal feed producer, knows all too well because of the high number of rush orders it receives. As a service to its customers, ForFarmers – together with the Mobile Innovation Lab – set out to develop a user-friendly, easy-to-use ordering application that advises on order quantity and suggests a delivery date. Feed-as-a-service, so to speak!

Meet Farmer John

Let’s check in with dairy farmer John, who uses ForFarmers’ ordering app. He is executing his daily tasks on the farm when his phone buzzes. It’s a push notification from the ForFarmers app: “It’s time to place a feed order.”

When John opens the app, he sees a personalized order proposal, including the volume and composition of the feed and a proposed delivery date. John checks the order carefully and wants to decrease the order quantity and delay the delivery by a couple of days. The simulation model in the app immediately calculates how much stock John has left when he postpones the order and signals that there is a problem. If John pushes back the delivery date, he might run out of one of his products.

John decides to revert the order and the delivery date back to the original proposal and places his order. While he gets back to his original daily tasks, the order is sent to ForFarmers, which immediately starts processing it. Four days later, the order is delivered to John’s farm.

The power of predictive analytics

ForFarmers’ app is groundbreaking because it predicts when the farmer needs to place an order, rather than measuring how much feed is left in stock. It not only predicts when the farmer needs to order but also how much of which products.

The app is able to do this by incorporating different datasets: the farmer’s order history, his livestock – the particular breed and the number of animals at the farm – and the weather forecast. The weather can affect how much the animals eat: in cold weather, the animals might need additional food, whereas they tend to eat less in hot weather.

All these parameters increase the predictive model’s accuracy. This enables the farmer to order as late as possible so that he can order as much as possible – and enjoy volume discounts – while ForFarmers always has plenty of time to process and deliver all orders. In addition, farmers using the app will get much better insights into their use of animal feed and ordering patterns.

Watch our video to learn more.

Twan van den Broek

About Twan van den Broek

Twan van den Broek is Customer Innovation Architect at SAP. Twan combines a broad experience in SAP development, Integration and Process Management with agile methodologies (Scrum) and Design Thinking to deliver SAP solutions that add value to your business and make end-users happy. As a crew member of SAP’s Mobile Innovation Lab, he delivers customer innovations with Rapid Prototyping within one week.