Use Case



Development and verification of an AI system for sales forecasting with forecast swing.


Mr. Toshihiko Adachi

Manager, Business Standardization and Business System Development Section, Business Promotion Department, Domestic Business Strategy Division, Domestic Business Headquarters, Omron Healthcare Corporation

Mr. Takashi Yamatani

Global IT Innovation Department, Management Control Division, OMRON HEALTHCARE Corporation

Background and Issues

The penetration of AI technology, the new coronavirus, and various other changes in social conditions have led to a reevaluation of the way business is conducted in the industrial world, and the shift to DX is rapidly advancing in many companies. In particular, in the supply chain, where a wide range of parties involved are intricately linked, DX is urgently needed because decision-making in planning operations, such as order volume forecasting and product shipment planning, is based on the experience of the person in charge. OMRON HEALTHCARE Corporation (hereinafter referred to as "OMRON HEALTHCARE") also used to have its staff members make forecasts of future order quantities, but due to the supply chain disruption caused by the Corona disaster and supply constraints caused by semiconductor shortages, the evolution of planning operations that appropriately capture demand fluctuations has become an issue.

About the installed system

In response to these challenges, OMRON HEALTHCARE aims to promote digitalization by developing an "order forecasting AI system that takes into account forecasted fluctuation ranges" using ReNom Plan, an AI developed by Grid, in order to build a robust and flexible supply chain that can appropriately capture demand fluctuations even under constrained conditions and allocate limited materials to appropriate demand based on demand fluctuations. The company aims to promote digitalization through the development of an "order forecasting AI system that takes into account the forecasted swing range" using the AI "ReNom Plan" developed by Grid.

Post-introduction effects

In the demonstration test, the average error rate in forecasting order quantities for major products was improved compared to conventional operations, which is expected to lead to future advances in the planning process, such as reducing the time required for planning and curbing excess inventory and shortages of products.