With very few exceptions, almost all S&OP systems today are “analog” which means much interaction and manipulation is needed from the planners to produce a working plan. There are two main reasons for the planners to be doing all the heavy-lifting: the system cannot represent the complexity of the supply chain (much like a spreadsheet) or inadequate availability of data. The latter happens because the data is assumed to be static and the system does not adapt itself to the on-going changes in the supply chain with millions of “moving parts.” Lead-times, suppliers’ performance, product mix, capacities, seasonal changes are changing all the time. A pre-defined value in the model causes inaccuracy in representation of the model. Model accuracy also has a huge impact on representing the physical supply chain. In a digital supply chain the model has to be a very accurate representation of the world with all the on-going internal and external changes. We refer to this as a Digital Mirror©, and Gartner calls it digital twin of the supply chain. Your typical S&OP, with a supply side spreadsheet based engine*, gives you a very long term but inaccurate visibility which is fine for stage 3 maturity but to go to a truly digital supply chain, the systems not only need to have accurate representation but it also needs to be self-improving as the supply chain changes.
A digital supply chain does not rely on pre-defined methods of predicting but adapts itself to changes. There are two kinds of changes: trends detected in the past and expected trends in the future. Examples of the past trends are equipment/tool efficiency, supplier lead-times and impact of weather on certain orders. As a result, a digital supply chain system needs to be an integral part of the whole IT echo system. It needs data from shop floor to examine availability/efficiency of equipment and people, it needs data from new product introduction to see how the product have launched based on their attributes and it needs to integrate the data from ERP to examine purchasing and supplier information, pricing changes etc. Given the availability of all such data, the system can identify patterns that are impacting delivery, cost, customer service, inventory points and so on and make corrections accordingly. The planning engine is constantly running and “simulating” or predicting the future. Hence it can highlight issues that keep coming up. After a while the system can deduce what are the main causes of potential latenesses, inventory shortage/excess and cost increases for use of substitute material amongst others. With such level of accuracy and adaptability, the system becomes touchless producing plans which are highly accurate and adaptive over time as the supply chain changes. We refer to this as self-improving supply chains©. To learn more on this topic and other AI related supply chain innovations click HERE.
*Almost all S&OP systems model the supply side using spreadsheet technology, i.e. fixed lead-times, bucketed capacities, pre-defined bottlenecks.