What Is Supply Chain Network Design?
Supply chain network design (SCND) can be described as the process that determines the best mix of suppliers, locations, logistics and production facilities for optimizing product manufacture and distribution. SCND entails answering questions such as where to source raw materials, how to transport products, what's the optimal location for production lines, product storage warehousing, and the best methods for delivering goods to customers. It refers to a process whereby a mathematical model of the supply network is created and optimized using various types of optimization software such as algorithms, heuristics, and linear programming.
Key Factors to Consider in Supply Chain Network Design
It's clear that supply chain network design represents a major departure from convention supply chain planning which usually focuses on factors such as the lowest product cost and just-in-time delivery. Because it analyzes and impacts the organization's entire supply chain, from start to finish, SCND should be tackled in a professional and organized manner. Many factors need to be considered, such as goals, objectives, costs, buy-in and the best way to model the organization to closely reflect actual performance. There are at least nine crucial factors to consider.
1. Determine Goals and Objectives and Create Buy-In
It's essential to have a clear idea of what the organization hopes to achieve as well as an honest assessment of the time, effort and cost. Supply chain network design is not an exercise that can be conducted in isolation; it must involve every function involved in the procurement, manufacture, sales and delivery of goods. It's absolutely essential for there to be high-level commitment from the CEO as well as the board. As part of this process, a pilot project that considers an element or part of the organization may be needed to demonstrate the benefits of supply chain network design.
2. Putting Together a Competent Supply Chain Network Design Team
Apart from high-level support, the success of SCND depends on appointing a team that has the requisite knowledge, skills and determination to see the project through. At the very least, there needs to be a full-time project leader and a small project team composed of individuals who, collectively, thoroughly understand the supply chain, the organization and a vision of what they want to achieve. Key executives and supply chain personnel must also be involved as and when required.
3. Selecting the Appropriate Modeling Tool and Vendor
As relatively few organizations have in-house experts with requisite modeling skills, it's likely they will need external consultants or modeling software vendors. Various modeling software solutions exist, including packages and platforms. Packages usually run in the cloud. Some are industry specific and can be used with relatively little preparation, but they may not effectively model all supply chains. Modeling platforms are powerful and flexible but generally require advanced coding skills. Some use drag-and-drop modeling, which is easier to configure and simplifies mathematical modeling.
4. Formally Defining the Existing Supply Chain Network
The first task is to define the existing supply chain network in some detail. This is an important, if relatively time-consuming, process. The level of detail required varies depending upon the type of analysis, but it's better to work with the greatest level of detail possible, as this means there are fewer errors. Typical information includes item, supplier, location, cost, distance, shipping method and constraints such as capacity, lead time and flexibility. Start with the bill of materials. A great deal of the required information will be available in the organization's MRP and ERP systems.
5. Modeling the Supply Chain
This information is fed into the optimization modeling software, creating a detailed mathematical model of the organization's supply chain. It is here that the benefits of drag-and-drop modeling are seen, as it takes less time to build the model and much of the work can be performed by employees without formal OR training. A team approach helps expedite model preparation.
6. Populating the Supply Chain Network Model
The third major task after defining and modeling the supply chain is data preparation. In most organizations, there's a wealth of data. However, additional sources of data may need to be identified and acquired, especially relating to forecasts and future market trends. These data streams need to be cleaned and prepared for use via various filtering techniques. Missing data is as important as faulty data. Once the data is prepared, it can be loaded into the model.
7. Model Validation
The model needs to validated. Irrespective of how it is prepared, certain aspects of the model will not be easily visible, so it's best to apply historical scenarios and compare the model output with a known result. There are two important aspects to model validations. For a known set of inputs, the model must be able to predict what will happen. Additionally, it needs to be calibrated to arrive at similar values in terms of financial figures and volumes. This important step generates confidence and credibility in the model.
8. Running Scenarios and Identifying Decisions
It's at this point that the model can be used to guide decision-making and to answer what-if scenarios. The model itself should simply replicate what happens in the real world. The next step is to use solver and optimization programs that can run multiple scenarios to determine the best options and decisions. For example, it would be possible to determine the best facility for manufacturing a certain product. External data streams, together with in-house forecasts, can help determine future strategic directions. An example would be whether to source products from countries in the Far East or alternatively to source similar products from closer but possibly more expensive suppliers.
9. Transferring Skills and Expertise
Supply Chain Network Design should be ongoing; it's not a static once-off exercise but rather one that is in continuous use. For this reason, it's absolutely essential organizations internalize the skills and knowledge to further develop and manipulate supply chain network models. This applies equally, irrespective of the type of model. If models are hard coded using a modeling language, competence needs to be transferred and institutionalized so the organization isn't dependent on one or two individuals who may (and will at some time) move on. In the same way, ongoing reliance on vendor support is expensive.
Key Takeaways for Supply Chain Network Design
Supply chain network design is an excellent technique for optimizing and managing an organization's supply chain. The modeling approach allows the analysis of different scenarios as well as helping determine the best supply chain decisions. While SCND is relatively costly, the benefits far outweigh the initial expense. Mathematical coding models are powerful, but they depend on high-level OR skills while drag-and-drop modeling offers greater visibility that's more easily internalized. OR and IT support is essential, as is the involvement of line managers and personnel involved in supply chain decisions.