Go Beyond Network Design

Network design is enjoying a high profile of late because it has become the capability that most companies with mature supply chain practices use to understand their costs, improve performance and plan. Once the practice of using network design is embedded in a company’s process it is used more frequently for making informed decisions. The realization that network design is essential for business success has taken many years to develop even though the technology has been available at a reasonable cost on desktops for over twenty years.

Network design is relatively simple compared to some other optimization challenges, but it still takes some training to understand the process needed to collect, clean and validate the necessary data. This data, including facility and customer locations, potential locations, production costs, warehouse costs, transportation details, taxes, customer demand etc., usually spans multiple systems.

In addition, network design requires some skill to understand how the optimization behaves and the important drivers. The results are only as accurate as the data, but this is usually sufficient, though not absolute, because supply chain costs are always flat around the optimum. This means that there are a range of options with similar cost results, not just one “perfect” optimized solution. Therefore, you can select the alternatives that provide other benefits such as flexibility, sustainability and localization without impacting the bottom line.

Over the years, network design has evolved from predominantly locating warehouses to locating plants, dealing with inventory, production planning, and transportation questions such as

  • Where and when should I start producing a product?
  • What is the most efficient mode of transportation between my warehouses?
  • What is the impact of warehouse size on inventory?

Often to answer these questions in depth you may need more detailed analysis using tools such as Transportation Routing, Multi-Echelon Inventory Optimization, and Network Risk. In addition, the customer demand information, typically derived from demand planning systems, can now be improved through predictive analytics that include using competitive data and machine learning for better results.

It is time to go beyond network design with a versatile platform that enables easy integration of all the components required and the ability to both model and prototype, and later integrate into an operational environment.

Strategic modeling requires different capabilities from operational planning and execution. The strategic focus is on generating scenarios, visualizing the data and results, comparing impacts, and collaborating with other users and decision-makers. The operations piece requires integration with current systems as well as a flow that will push results back to a system of record. Therefore, the operational focus includes ingesting data from various sources and integrating prebuilt solvers and data wizards into a transparent flow for business users.  An ability to create, edit and visualize the workflow and load the results to the right systems is essential.

The Opalytics Cloud Platform has been built from the ground up to deliver all of this. It provides a way to both model network design and build an operating model that includes other solvers and integrates with existing systems. This is precisely what we have done for a large retailer around production planning and forecasting as described here. The result is a seasonal production plan with seamless data flows among forecasting, planning, and client operating systems, including ERP.