Production planning for a diary supplier
One of our clients, a leading Belgian manufacturer and supplier of high-quality dairy products, contacted our team to help them with production planning.
The production planning of our client takes place weekly. It is very challenging because of the rich portfolio of products that our client produces and the uncertainties associated with the supply of raw materials and purchase of end products.
One of the biggest challenges for supply managers is handling volume demand volatility. Demand forecasting tries to estimate future orders or requests to pre-emptively detect resource and capacity shortages. Customer interactions and expectations are becoming increasingly complex due to more diverse and more personalized products and services entering the market.
At our client, raw milk is delivered to the site and further processed. In this process, among other things, the fat is extracted from the milk in which case skimmed milk and the fats are obtained as two basic ingredients. Afterwards, the two basic ingredients are recombined in a ratio that depends on the end product to be made. We sum four important features of the planning:
- Freshness and storage restrictions – the client produces natural dairy products which means it can only store these products for a certain length of time and there are also limitations on the quantities it can store while maintaining freshness.
- Pack size and production planning – the client can produce as many small bottles per hour as it can for big bottles, though it uses fewer raw ingredients for the smaller pack sizes.
- The change-over-time required to switch products – eg. After certain products are produced, the line must be shut down and cleaned before any other can be produced.
- The process of forecasting demand was a one-man job that has been performed for more than 15 years – The expertise he has built up over the years has never been written down or transferred to anyone else. This causes a serious risk when this person will be absent for a longer period of time or leave the company as all this knowledge and expertise would leave with him as well.
We tackled this problem in 3 stages:
- We looked at the existing demand forecasting to see where to integrate AI.
- We identified the data sources that we could extract, including past sales orders, and shipping and price lists. We then looked at how to feed this data into the model
- We pinpointed any anomalies in the data.
- We developed a solution by selecting the best forecasting algorithm that could predict the required weekly volumes for each product line until we had reached the business success criteria (accuracy >90%, weekly basis, split by vendor, etc.).
- We fine-tuned the final model, debugged it, and got it ready for production.
- We looked at where the model would be deployed, either in the cloud or on the company’s existing infrastructure. We also looked at feeding back results into the existing forecasting process.
- We created pipelines for the data to flow to the model.
- We resolved the issue of ‘cold starts’ where a new product appears that does not have historical data. We used data from comparable products to forecast demand for these new products. We designed the AI model in such a way that it could ask questions to humans to support this process.
How we added value
- Wasted end products reduced by 30%
- Ability to optimize production schemes since the weekly required volumes could be forecasted with very high accuracy
- Sales increased by 1% as in-demand products remained in stock
- Our client freed up staff (0.5 FTE per year) thanks to our automation of its manual forecasting process