Big Data Means Big Quality Data
One of the pillars of Industry 4.0 is availability of data so that we can process it intelligently. Allowing us to form conclusions and making new discoveries that were not so transparent to us in the past. In order to build smart factories and autonomous supply chains, the very first step is to build a digital model of the supply chain. to build such a model, good data is essential at high frequency. That is the more granular and more frequently the data is updated the better digital twin of the operations we have.
As pointed out in an article by McKinsey & Co., you need to know what data is critical to your operations and what is being used, or should be used, regularly. And figure out ways that can ensure accuracy and timely availability of data. Above cannot be done without a digital model guiding you to identify bad from good data and demand frequency of its availability. Systems can find inconsistencies in data and missing links. They also have little patience if data is not available when needed. They will make sure that you know about it!
The immediate question is, what data do I need and how frequently? Or do I already have the data? is it good enough? It is hard to answer these questions unless you know what you want to do with your data. In other words, the kind of data that is needed and the method by which it can be analyzed in order to get the results that we are after. Let’s say that you are interested in running your factory lights-out. This implies that every piece of equipment needs to be modeled digitally. You also need to have real-time availability of WIP data and awareness of the order requirements in detail as to exactly how it can be made, alternative ways of making it and what to do in case of equipment breakdown or quality issues. Thus, we are one step closer to knowing what data is needed and how frequently. We need static data to build the model or digital twin of the factory or supply chain. We also require dynamic data as frequently as possible to update the model accordingly. Therefore, processing times, tool requirements, batching capacities, changeover times, substitute materials, alternative routings etc. are all essential to build the model. Then we need to have connectivity to operations. perhaps through an MES system, IIOT or streaming data to know and measure our progress. Finally, in order to take the right actions, intelligence is needed to respond and predict events to optimize and mitigate risk.
Supply chain planning systems such as S&OP and S&OE are intended to give you visibility by building a digital model of your environment and guide you to have correct data and enough intelligence to optimize operations. Obviously, S&OE system can do a much better job because they demand higher granularity and frequency of data. They are much closer to a digital twin of the operations than the high level and long-term models of S&OP systems.
To learn more about how systems can intelligently run your supply chain operations and factories autonomously; and what data is needed to get there, visit Adexa.
