Proactive inventory management that reads future changes in demand is key
The term “supply chain” refers to the series of operations involved in a product production workflow, from raw materials to delivery of the final product to the customer, in tandem with various services. Today, optimizing supply chains through ICT, constructing appropriate production systems, reducing excess inventory, and ensuring fast shipping systems— in other words, accurate supply chain management (SCM)— is key to improving profits in manufacturing and distribution.
In SCM, special emphasis is placed on inventory management. For example, makers who manufacture home appliance products as well as mass marketers and other distributors who sell such products think hard about how to adjust inventory and maintain it at an appropriate level. If sufficient inventory is not secured, products cannot be provided on a timely basis. On the other hand, storage and management of products requires time, effort, and money, which may place added pressure on management. Ideally we want accurate demand forecasts identifying precise amounts of how much a product will sell, however demand often fluctuates due to sales promotions, the introduction of new products and what competitors are doing in the market. Highly accurate predictions that take into account such fluctuations are difficult. For this reason, quick decision making about how raw material orders should be placed and to what extent production volume should be increased is an ongoing challenge for many organizations.
Minimizing total costs to improve retail profits
Recently, tools that enable the collection, accumulation, and analysis of massive amounts of product information from social media in real-time— in addition to product sales data through the use of POS data— are rapidly being developed. Even retailers, who experience changes in demand due to uncertain factors such as sales promotions and new products, have strong expectations for highly accurate demand forecasting using massive data (big data). This will effectively lead to better efficiency in decision making opposed to the conventional seat-of-the-pants approach.
Fujitsu has developed model predictive control technology specifically intended for supply chain management. This technology reads future changes based on multiple long-term forecasting scenarios for industry sectors where sales vary due to uncertain factors. The technology is used to develop ordering and production plans that maximize long-term profits, thereby minimizing total costs while making successive revisions to plans developed in advance. With this technology, it is possible to develop highly accurate plans for responding to sudden changes in demand and ensure optimal decision making that takes into account risks. This is achieved by considering inventory holding costs and lost opportunities due to running out of stock as well as making revisions to the predictive model. When confirmed using actual store data for approximately 90 retail stores over a period of 60 weeks to generate ordering plans, profits at all stores improved compared to the conventional method. In total, this technology increased inventory management-related profits by an average of 16% (approx.).
Fujitsu aims to develop solutions that assist in operational and management decision making by combining advanced forecasting and analysis technologies based on big data.