Predictive Performance & Manufacturing
December 20, 2017
Industry 4.0 at times gets sidelined or lost under heavy words like augmented-reality, artificial-intelligence, big-data/advanced analytics and so on. But it is a fact that this modern revolution in manufacturing is centered in data management and use of analytical tools to predict business metrics - all this to enable better and more informed decision making.
As we have seen in previous posts manufacturing IT applications, like the MES are to play a key role if I 4.0 is to be fully manifested in an organization’s supply chain. Today, we will focus on how the MES coupled with other important technological advancements related to I 4.0, not only improves process performance monitoring but allows, the process performance to also become predictive in nature.
Say there is a manufacturing plant which assembles electronic equipment, for instance a mobile device manufacturing plant. Let’s also assume that the operation of this plant and members of its value chain are monitored/controlled/managed through an I 4.0 ready MES, at the center of the IT infrastructure. Now since the process in main manufacturing plant and the value chain is connected to the MES, it can orchestrate the entire process and provide better clarity to process owners and top management.
Is your MES Industry 4.0 ready?
Manufacturing Software for Industry 4.0. Embracing Change and Decentralization for Success.
by Iyno Advisors and Critical Manufacturing
This White Paper defines a path to Industry 4.0, focusing on the following aspects:
|I4.0 concept: enabling technologies, cyber-physical systems, etc.|
|role of MES in Industry 4.0,|
|steps companies can take now to prepare for Industry 4.0.|
Imagine a scenario, where a particular dealer for the company’s mobile phones reports faulty casing for a batch of phones received at their company’s warehouse, the defect rate reported in more than 2%. As soon as this error is reported, the MES kicks into action, acting simultaneously at multiple fronts.
Immediately an intimation is sent to the supplier of the phone casing to, firstly arrange replacements, secondly investigate their quality related data, to understand how an error was missed. Internally the MES immediately deploys advanced data analytics tools to understand how such an error escaped through various processes of the main assembly process.
If it is detected that metrics set were not adequate to detect such an error, process owners would receive a report allowing for them to change minimum standards. At the same time the MES would also revise KPI and process metrics to show effects of complaint registered and initiate dynamic scheduling tools to ensure supplier is provided correct mobile phones to fulfill the order, this is still the containment part.
Predictive modeling - seeing the future
Now moving on to the wow factor, based on the information received with the MES, and prior data available with the MES, the application would also provide process owners with predictive models. These would not just be related to the particular error reported, but concerning the overall throughput, OEE, TCO and other key metrics.
Modern MES applications have the ability to use even the most mundane looking production data and provide predictive models related to process performance. In above case imagine that the main manufacturer of the casing discovers that one of their molds has a defect. The MES could actually through analytical models, depict exactly when the mold started to have issues, where the product was still acceptable but moving towards unacceptable and also the future projected loss if the manufacturer would continue to employ that particular mold.
Having such valuable information would definitely lead to better decision making, thereby allowing all value chain partners to benefit from the effectiveness of I 4.0.
Data analytics and statistical tools have been around since a long time, however, what the MES and I 4.0 change is the ability to detect patterns and trends from transactional data recorded every single second from the shop-floors of entire value chain manufacturers. The MES can predict decline in OEE, increase in TCO, reduction/increase in throughputs, based on present and past data, but unlike manually applied tools, it is automatic and real-time, which allows process related improvements to be implemented immediately.