According to Gartner, there are a number of factors that are essential to enable a digitized plan. At Adexa, we have designed a digital Blueprint that addresses the issues discussed below so that companies can take incremental steps to get to the highest level of maturity (Level 5 according to Gartner) by plugging in technological innovations that are needed. The Digital Blueprint© is designed to overcome factors that hinder generation of accurate plans such as dependency on manual planning, removal of subjective planning and bias, inaccurate models and data quality issues on an on-going and dynamic basis. These factors are:
Demand signals and changes in forecasts as well as changes in operations such as yield factors and equipment availability, cause inaccuracies in developing an accurate plan. Although many sources of variability are outside of our control, however, measures can be taken to mitigate the impact. An example of this is Multi-Echelon Inventory Optimization. It is a prescriptive analytic technique that can closely predict what levels of inventory to be kept at different stages of a supply chain to gain the desired service level at the lowest cost. In demand planning, predictive techniques are also available to pick the best policies that are known to give good results based on product attributes.
- Model Accuracy
This is probably one of the most important elements of a digitized supply chain. Unless you have an accurate model of your operations, suppliers, products, any result produced by the system is not very dependable and requires a lot of manipulation by the users in order to make it work. Spreadsheets are a good example of poor quality models when used for planning. They assume fixed lead-times, bucketed capacities, and do not take product mix into account. Needless to say that almost all S&OP solutions use the same technique for supply planning! On the other hand S&OE, forms an accurate digital blueprint of the supply chain taking into account accurate capacities, tools, processing times, alternate methods of manufacturing and delivery and so on. With Sales and Operations Execution S&OE plans are 98% accurate and executable without any manual intervention.
- Data Quality
Unless there is good data available there is no chance of getting a true digital supply chain. Generating good plans depends on timely and up-to-date as well as complete data. Typical data that is needed can be divided into two groups of static and dynamic. Amongst the former group one can include routings, sourcing methods, and BoMs. Dynamic data, depending on the sophistication of the system, can be real-time MES data from the shop floor, material availability and demand sensing as well as ATP/CTP capability. To this end integration with other systems is important to support the real-time operation of the digital supply chain with MES, ERP as well as PLM systems for NPI. Furthermore, as part of the digitization, we need to ensure the availability of adequate MDM system for data correction and validation.
Users tend to have their own biases leading to subjective decisions which are intended to do the right thing for the company. However, with all their good intentions it is quite possible to create unintended consequences that can cause sub-optimal results. Systems can deal with millions of data items to arrive at a decision which is optimal or near-optimal, but they do not have human intuition (until now!) On the other hand, people can only handle a handful of variables but with experience and intuition make decisions that are good enough but not necessarily optimal. In other words, they are not aware of a potentially better solution that can support, perhaps significantly, the objectives of the company. Biases come in different forms, such as forecasting, production volumes, and meeting one customer’s need at the cost of many other customers’ service level. Automation removes bias and can lead to an objective and optimal decisions at a much faster pace.
- Self-Improving Supply Chains
Supply chains are living organisms, they change, they grow and shrink and the world around them is changing all the time. It happens at a slow pace that may not be noticeable until it is too late. For example, supplier delivery patterns and lead-times can change over time, yield factors can improve, or machine capacities can decrease the longer they are used. Relying on old data to reflect the operations can result in an inaccurate outcome. A Digital Blueprint enables machine learning techniques to constantly monitor the underlying changes in the supply chain, past, and future, and make the corrections accordingly. Hence having an up-to-date model of the operations dynamically.