Case Study: Forecast and Production Planning

Network optimization has traditionally been used to make decisions about where to locate plants and warehouses, how to best ship products and where to source. This has expanded into decisions around production planning at a high level. These include which plant should produce which product and when, given capacity and cost constraints.

This type of application makes network optimization an essential part of the Sales and Operations Planning process where it can be used to improve production performance based on demand. This analysis is also an important part of long term risk assessments and short term problem resolution.

The following is a case study of a company that recently deployed Opalytics applications for this type of analysis with a high return on investment.

Company: A global retail company serving thousands of stores in over 20 countries.

Background: The Company produces thousands of  SKUs in hundreds of  manufacturing plants located in different countries in Asia and ship the products to over 50 distribution centers all over the world. Every year, production is planned for four different seasons: fall, spring, summer, and winter. Since capacity is limited and lead times are long, sometimes over 8 weeks, the company needs to pre-build inventory before the season starts as well as coordinate production and distribution as follows across geography (which plant should produce what product), across time (when should a plant produce a specific product for a specific season) and across channels (which plant should serve which distribution center).

The decisions include:

  • When to start production for a specific SKU and which plants should produce that SKU
  • How long to produce a specific SKU and how to pre-build inventory
  • Which plant should serve which distribution center.

The objective is to plan production so as to minimize total production costs (different plants may have different costs for the same product), inventory holding cost and duties as well as and transportation cost from plants to DCs.

In addition, they take into account other constraints including balancing production utilization across time on every plant and every production line and satisfying lead time requirement constraints by product. For instance, delay production as much as possible for all products whose forecast accuracy is low.

Outcome:  

The company deployed Opalytics for forecasting and production planning, Before the Opalytics deployment, decisions were made manually by the production planners. Now they apply Opalytics Predictive Analytics for generating seasonal forecasts and Opalytics Network Optimization to allocate and optimize production and distribution. This is the only way to find, among all possible production and distribution strategies, the one with the minimum total cost that will satisfy all the business constraints.

The impact was dramatic! Total costs – focusing on manufacturing, duties and transportation cost – were reduced by 6%, response time improved by one week and some production was reallocated to lower cost locations.