Supply chain network optimization seeks to find an optimal combination of factories and distribution centers in the supply chain. The solution should match supply and demand, as well as find a network configuration with the lowest costs. Based on the optimization results, a manager can compare potential network designs and evaluate the maximum profitability of each of them.
Supply Chain Network before and after optimization in anyLogistix software
In short, a supply chain model can include a set of all possible flows and facilities. To get the most efficient one, you can perform a network optimization experiment in anyLogistix — a tool for supply chain design, optimization, and simulation. The output data will show values of transportation and production flows, inventory at the end of each time period, and associated costs. Furthermore, you will have data on several possible network configurations options, including those with the lowest costs. From that, you can then choose the one that best fits your business and implement it in the real world.
Learn how to use supply chain network optimization to find optimal DC locations.
In real life, companies may have specific requirements that are essential for their businesses. These requirements should be taken into account in the supply chain network optimization as constraints. Constraints could be, for example, a limited amount of facilities, or step costs (fixed within certain boundaries).
In anyLogistix, constraints help the software decide which solutions would be feasible. When setting the parameters for the network optimization experiment, you can choose values for:
However, it is not always possible to get an optimal result that complies with all of the requirements. If the strict operation rules you’ve set for the supply chain are controversial, no solver will be able to find a solution. In this case, think about changing some of the hard constraints into soft. This means that you allow disregarding some of the rules you have set. For those violations, anyLogistix would charge penalties while calculating an optimal supply chain result. This provides you with a supply chain network optimization solution and insight as to why some of the constraints could not be satisfied.
As every business is unique, every supply chain has specific requirements that can’t always have predetermined values. For this purpose, anyLogistix has custom constraints. You can describe them with variables, equations, and conditions. Custom constraints are applicable in many cases. For example, when you need to maintain a certain product flow ratio between suppliers, or determine a safety stock percentage from the incoming flow that a distribution center should build, to name a few.
In case you haven’t set an objective for your supply chain network optimization, anyLogistix provides the best solution based on maximum profit.
Watch the webinar to learn more about constraint-based network optimization in anyLogistix >>
Master planning is an extension of the supply chain network optimization technique that helps synchronize production, storage, and transport with demand. In short, it answers the questions: How much and where to store? How much and where to produce?
For master planning in anyLogistix, a scenario is split by periods (e.g. weeks, months) that relate to each other. For example, one period’s output is another period’s input. Additionally, for each period, there is a certain demand value. During the experiment, the optimizer calculates, for each period, if a facility should be closed or open.
Read more about master planning >>
To perform the anyLogistix Network Optimization experiment, you need to set the demand for the products in your supply chain as well as the locations of the suppliers, customers, and facilities (warehouses and DC’s). If you don’t have such input data, you can generate it with the Greenfield Analysis experiment (GFA).
For Network Optimization, you might also consider different types of costs (transportation, facility-associated, etc.), time periods, and most importantly, constraints. After setting the necessary parameters in anyLogistix, you run the experiment and the software optimizes your model with the built-in solver IBM ILOG CPLEX®. The result is an optimal solution — a set of facilities chosen according to:
What’s more, the user can further change or add the optimization model parameters and run the experiment again to get the result that complies with all of the requirements.
After you’ve tried several network optimization iterations, for a deeper what-if analysis and risk estimation, you may want to convert the results into a simulation model scenario.
You can use network optimization for several supply chain design problems:
See how various companies used anyLogistix for supply chain network design and optimization:
