How to evaluate process and quality data to boost your production

As a Solution Architect, the author Henryk Fischer works on creating new Industry 4.0 solutions close to the customer and is responsible for the Unide open source project at the Eclipse Foundation. He was also involved in setting up the customer project with Promess and Bosch, described in the use case below.

How can you judge the quality of a process and take steps to improve it ? The first step is to get a thorough understanding of the process. But I think this shouldn’t be something new for you. However, getting this understanding can be a real challenge sometimes – especially when working with closed processes that are highly complex. Think about the production of manufacturing parts like spark plugs, windshield wipers, or high-pressure pumps! Is there an easy way to create transparency within these production processes?

Yes, there is: Dedicated software for monitoring process quality increases the transparency of the production process . The software makes it possible to digitally map and analyze manufacturing processes and take appropriate measures for boosting product quality.

A standardized machine language for process and quality data

To analyze this wide range of different manufacturing processes you need a common “language” for the process and quality data. This common format needs to be independent both of the process in question and of the software program used to perform the analysis. The great news is: Such a format already exists! The Production Performance Management Protocol already offers a standardized format for measurement values and machine feedback . The Eclipse Unide community, that created the protocol, expanded it now to include a new type of notification: the process message.

How do I use it in practice ?

Now, the Production Performance Management Protocol makes it possible to summarize processes in a single message, which the respective Industry 4.0 solution can then work with further. The message abstracts the actual process, e.g. pressing, welding, or tightening, regardless of type.

But which kinds of data does a process message include so that the software can then evaluate as needed? Let’s take the example of a pressing process:

  • For one thing, the process message contains information about the device that carries out the process. In this example, this could be information about the press.
  • It also contains details about the process itself. This can be program data or shut-off values that determine when the press needs to terminate the process.
  • And it delivers information on the finished parts or batches, for example, the ID of a metal part, that has been pressed.

This links the process data to the quality data and ensures a better understanding of the evaluated processes as well as of the correlations between process parameters and product quality.

Infographic shoing how the PPMP works.

Structure of the Production Performance Management Protocol process message.

What are measurement values and special values?
Measurement values describe the process sequence. They, in turn, are split into phases characterized by special values; for example, values for how long to maintain a certain level of force, or extremes in a defined range. Based on this data, the software can quickly evaluate process quality without having to analyze the entire process curve.

How Bosch & Promess are visualizing and evaluating ECU pressing processes

The Bosch plant in Ansbach makes electronic control units (ECUs) for airbag, ABS, and ESP systems. Mechatronic presses made by manufacturer Promess assemble the individual components of these ECUs. With the help of the Production Performance Management Protocol data is extracted from the proprietary control units of the presses and sent directly to a software that is evaluating the process and quality data. At Ansbach plant, they use the Production Performance Manager for this evaluation.

Screenshot Production Performance Manager

Evaluation of pressing processes at the plant in Ansbach. In the software, you can click on the green and red bars at the top and get a visualization of the force/position curves for every single process.

Promess is one of the first machine manufacturers to have independently implemented the Production Performance Management Protocol for process and quality data. This saves the company the time and cost of developing special gateways or integrators for extracting process data for the end user (the plant). The Promess presses now send their data directly to the appropriate software.

Why does the evaluated process data not need a common point of reference, such as time?
Time references are not strictly necessary for visualizing this data. In contrast, continuous measurement values are always captured in relation to time (e.g. temperature progression of cooling water). As a result, users can correlate non-time-based values (for example, force to position).

Based on the data from the control units of the presses the ideal form of each process is defined and thus serves as a reference for each and every pressing operation in the plant. The software enables users to evaluate each process directly and immediately based on the complete set of raw process data. Previously, this was possible only by random sampling in the downstream quality assurance phase. Furthermore, the resulting transparency of historical data makes it possible to identify parameters that are critical to the process but have thus far gone undetected.

Summary

More from Henryk Fischer

Video: Get an easy explanation of the Production Performance Management Protocol.

Listen to Henryk introducing Eclipse Unide – A way to establish an open Industry 4.0 standard in this webcast from the Eclipse virtual IoT meetup.

Looking for something more hands-on? In this video Henryk gives you an overview on what it’s like to hack with the Production Performance Manager.

 

About the author

Henryk Fischer

Henryk Fischer

I have been working at Bosch Software Innovations since 2016. As a Solution Architect, I work on creating new Industry 4.0 solutions close to the customer. Another responsibility of mine is the Unide open source project at the Eclipse Foundation. I joined the German Army in 2004. After I became an officer, I studied mechanical engineering with a specialization in artificial intelligence and automation technologies in Hamburg. Following that, I became a helicopter pilot in the Army Aviation Corps. In 2012, I left the Army and worked as a product owner and team leader for a green energy supplier in Hamburg.