Agentic AI saves L’Oréal 60,000 hours a year

Many companies are exploring the possibilities of agentic AI in their supply chains, but what does this technology actually deliver? At Manhattan’s annual conference in Milan, L’Oréal outlined the savings achieved using AI agents for wave planning, replenishment and cycle counting. The cosmetics manufacturer also explained the shift in approach that led to success: ‘‘We first wanted to measure the actual value delivered by the agents.’’

Wave planning is the process whereby the warehouse management system (WMS) selects a group of orders and releases them for order picking. Before the WMS proceeds to do so, the system checks whether the required stock is actually available. However, problems can arise during order picking, for example if the stock is administratively available but has not yet been moved to the picking location. If it becomes apparent on the warehouse floor that an order cannot be picked as a result, the WMS deselects it and removes it from the wave.

Cosmetics manufacturer L’Oréal has implemented two AI agents over the past year to help prevent these problems. The ‘pre-wave agent’ attempts to predict in advance which errors might occur during order picking and proposes solutions. The ‘wave coordinator agent’ has been developed to analyse and resolve problems that still arise despite the preparations. ‘‘With success,’’ reports Łukasz Kierzkowski of L’Oréal. ‘‘In one of our best-performing distribution centres, the number of rejected orders has fallen by 80 per cent in three months. The productivity of the order pickers has risen by 14 per cent.’’

This story was originally published by Supply Chain Movement. Read the full story at Supply Chain Movement →