The Real Value of Analytics in Industry 4.0

We constantly hear buzzwords such as Industry 4.0, IIoT, big data, machine learning, digital transformation and many more, but what real value do these concepts offer the manufacturing industry? Let’s start by defining some of these concepts.

Industry 4.0, also called the Fourth Industrial Revolution, encompasses all the previous concepts plus much more. It is the concept of a connected manufacturing environment that includes not only the machines and robots in an industrial space but also the people, processes and systems within it.

Think of Industry 4.0 as a long-term vision or strategy that typically requires a phased approach to realize. For example, before you can apply artificial intelligence algorithms to optimize your manufacturing processes, you need to capture relevant process data. So, capturing that data would be the first step toward realizing the real value of Industry 4.0.

McKinsey Global Institute observes that this Industrial Revolution drastically differs from the previous ones. To get value in the previous revolutions it was necessary to invest heavily in capital equipment. 80 to 90 percent of the value that manufacturing industries realized came from such investments. In this new Digital Revolution, the value comes from the data. As Geoffrey Moore pointed out: “Without data you are blind and deaf in the middle of a freeway.”

Even with all this value locked up in data, many of today’s machines remain unconnected. Industries face different challenges in utilizing this data, including old machines with legacy controls that “speak” different protocols, no standardization of industrial networks, different objectives between OT and IT departments, and security concerns. These challenges, however, should not discourage manufacturers from taking the first step down the path of digitalization since the ROI will easily overcome the investments. Typical first steps may include using “data collectors,” isolating cells and implementing policies that adhere to industrial standards.

The value generated from the application of this IIoT or connectivity will usually come in four forms: 1) Operational Efficiency; 2) Predictive and Preventive Maintenance; 3) Supply Chain Management and 4) Inventories and Logistics. The numbers are by no means insignificant — the results from implementations in today’s factories show that OEE (Overall Equipment Effectiveness) can improve up to 23 percent, productivity up to 30 percent, and inventory control up to 48 percent. These figures do not even include the application of exotic AI algorithms; they come from just applying common, standard analytics. Just gathering the data from the machines in a process or manufacturing cell allows manufacturers to detect areas for improvement.

Let’s consider OEE, one of the most useful metrics in production lines or industrial machinery due to the fact that it uncovers all inefficiencies of the process. OEE is represented in percentage, from 0 percent to 100 percent. An OEE of 100 percent is considered perfect production, meaning you are manufacturing without any defects, at full speed with no downtime in your production line. There is much literature out there to consider what is an ideal benchmark, but as a general rule of thumb and for convenience, you can consider the following: An OEE of 85 percent and above is considered world class. If you have achieved it, you should be proud of it and congratulate your team. However, you still need to consider that there is 15 percent of loss and you do not want to stop there, so a continuous improvement plan becomes convenient at this stage. Below 65 percent indicates that there is an urgent need to improve your process. You have a low competitiveness and a serious analysis is required to implement a plan to improve your OEE. Any score between 65 percent and 85 percent is considered acceptable but, again, you do not want to live with such a gap in losses and the continuous improvement philosophy plays a key role in becoming more competitive.

The formula to calculate OEE is OEE=A x P x Q where A=Availability and it is the measure of the run time that a manufacturing line or machine was planned to be running, taking into consideration any unplanned stoppages. P=Performance, which is the measure of the real run time versus the maximum run time that the process should be running, and finally Q=Quality, which considers all defects, including re-work. With all data in a single location, you are able to analyze all three variables, find the root causes that are reducing your OEE, and often gain substantial improvement within days of your initial implementation.

In summary, Industry 4.0 is a strategy — a journey that requires extensive planning and alignment of resources, processes and systems within your organization. As with any extensive project, it can start with a phased approach, and you can harvest some benefits even in the initial phases if you do it correctly. Small investments and small steps that do not require major surgery in your production lines can yield significant initial returns.