Which Company's Case Study?
This time, we introduce the case study of Unilever, one of the world's largest consumer goods manufacturers. Known for brands such as "Dove" and "Lipton" and operating in 190 countries worldwide, the company manages an extremely complex global supply chain. To streamline this vast network and respond quickly to market changes, Unilever is company-wide advancing the utilization of AI and digital twin technology.
Challenges They Wanted to Solve
The company's challenges were the complexity of its global supply chain and the inefficiencies that came with it. With frequent unpredictable events such as climate change, geopolitical risks, and sudden fluctuations in consumer demand, traditional planning methods were no longer sufficient. Specifically, they faced issues like demand forecasting accuracy, the associated risks of overstocking or stockouts, production plan optimization, and increasing logistics costs.
How AI Was Used
Unilever, in collaboration with partners such as Microsoft and Accenture, constructed a "digital twin" that replicates the entire supply chain in a virtual space. This digital twin faithfully reproduces the movements of the physical supply chain, from factory production lines to logistics networks and inventory status, using real-time data. AI platforms like Azure AI are integrated into this, incorporating and analyzing external data such as sales data, market trends, and even weather and traffic information. This enables highly accurate demand forecasting, simulation and optimization of production and logistics processes, and proactive detection of potential risks.
Implementation Effects and Key Takeaways
- ▸**Cost Reduction and Productivity Improvement**: In one factory, the introduction of digital twins led to a cost reduction of $2.8 million and a 3% increase in productivity. Furthermore, on the deodorant production line, they achieved 95% predictability for process issues, reduced waste by 20%, and increased production capacity by 10%.
- ▸**Inventory and Sales Opportunity Optimization**: By introducing AI-equipped smart freezers that automatically place replenishment orders based on real-time inventory data, they prevented stockouts in stores, achieving sales increases of 8% in Turkey, 12% in the US, and 30% in Denmark.
- ▸**Towards Data-Driven Decision Making**: The key point is not just the introduction of AI tools, but the construction of a foundation that integrates and visualizes data across the entire supply chain in real-time. This has enabled on-site personnel to make quick and accurate data-driven decisions.
- ▸**Achievement of Holistic Optimization**: It is important that they moved away from previously siloed, partially optimized processes. By simulating various scenarios on the digital twin, they can now derive optimal strategies from a holistic supply chain perspective.
What Japanese Companies Can Learn
Unilever's case study offers rich insights for many Japanese companies that manage the physical flow of goods, such as those in manufacturing and retail. While building a company-wide digital twin immediately might be challenging, it is possible to start small with initiatives like "demand forecasting for specific products" or "production line optimization for a particular factory." The important concept is to begin by integrating fragmented data and establishing a "single source of truth" that AI can analyze. This approach of using data to improve forecast accuracy and eliminate waste can be applied regardless of scale.