The team is also using machine learning to achieve right-first-time sealing of packs to increase product quality and reduce waste, and to predict the optimal time to carry out equipment maintenance to reduce maintenance costs – which has almost halved since 2018 – and maximise machine availability.
The ability to ship products direct from the factory is a real competitive advantage for the powders business in Brazil, given the complex customer distribution network. In the past, customer order allocation was prone to human error because of 600 daily decisions, 13 critical variables and constant changes. Using a machine-learning algorithm, the team can now predict the ideal allocation and route, using real-time data. This has reduced distribution costs while improving inventory and service levels.
At the heart of this transformation is a digital training programme which has upskilled the entire Indaiatuba workforce as well as more than 70 employees from seven other Unilever factories across the region. The team also established a tech ecosystem which engages more than 35 partners – including start-ups, universities and suppliers – for the fast prototyping of new solutions.
As Francisco Betti, Head of Advanced Manufacturing and Value Chain at WEF, says: “Lighthouses are demonstrating how to scale advanced technologies across entire manufacturing networks and beyond towards suppliers and customers or new functions such as procurement, logistics and research and development.”