Intelligence for Optimizing Production Lines, Part 1: Meaning and Purpose

Smart Manufacturing is being in headlines for Industry 4.0 applications, but the essence is often hidden behind tons of information. With this post I'm staring a series on Intelligence in Manufacturing. My goal is to explain and initiate a debate on what is happening in reality in the scope of production line optimization, and do that in a down-to-earth way. This first post touches upon the meaning behind the term intelligence when applicable to manufacturing, and more importantly - its purpose.

Meaning

Starting with what actually is Intelligence in Manufacturing. Simply put, production generates a lot of data - batch information, machine settings and sensor readings, quality control information - all that data can be processed and analyzed to identify actual or potential problems and get insights for improvements. The approach to process and analyze data can be helmed under the umbrella of intelligence.

Purpose

There is a notion about replacing human intelligence with the intelligence coming from machines. I regularly bump into that in conversations with customers' users, in professional community forums and in the media.

My take on this is largely influenced by experience and is straightforward:

The purpose of intelligence coming from machines is to enhance and support the human intelligence to achieve more

Intelligence in manufacturing can be seen as a tool for the stakeholders - most notably Process Engineers, Operations Managers, Operators. With advanced tools they can tackle the problems that they simply could not address without.

The analogy for that is evolution of tools for construction works - it's possible to build a one-family house completely by hand. To do more - to build an apartment building that can host multiple families, and do so effectively - construction specialists need more advanced tools - cranes.

Similarly, the intelligence in manufacturing is that crane that helps specialists running production process do so better and hugely extent what they can achieve.

With intelligent tools at their disposal, production process specialists can divert their attention to the completely new topics and continuously raise the bar of possible production improvements

I will illustrate that with a simple example. Let's imagine an operator who is monitoring the health of production line by looking at real-time data on temperature and power consumption. The job of the operator is to detect if something goes wrong and act to resolve it. The image below shows the approach without intelligence - plain data. Even just two parameters like that can take full attention of the operator in case the production process is fluid.

Now let's add simple intelligence - rules to highlight when Temperature goes below a threshold of 180C. The attention of the operator will be attracted by a red mark - that means that he or she does not need to constantly look at the charts and can extend the scope of their job and add more value.

What's Next

Firstly, debate. I'm interested in the community response - do you agree and notice similar trends? Or your point of view is slightly or completely different?

Secondly, in the next episode, I'm planning to go through the high-level types of intelligence that can support production line optimization, and illustrate that by examples.