Putting M4.0 to the Test

By digitizing and automating assembly and test processes, manufacturers can make strides toward improved quality and better efficiency.   

In discrete manufacturing, M4.0 initiatives often are challenged by the variety of processes used and parts made in assembly and test. Manufacturers can improve product quality and reduce production costs by digitizing and automating A&T processes.

An incremental implementation of these methods can provide quick wins, establish priorities for next steps, and lead to the discovery of more areas for improvement.

Connecting M4.0 with Assembly & Test

Jon Hobgood, VP of Advanced Manufacturing and Automation at Honeywell, says that M4.0 efforts should target three elements of manufacturing:

• Lean
• Digitization
• Automation

Lean initiatives have been in place for decades. Almost every discrete part manufacturer uses Lean at some level. However, digitization and automation are often not used at their A&T stations in ways that bring to bear the synergy of these three elements. Reluctance is based on expectations that implementation is too costly and solutions are years away. However, incremental steps are achievable, and benefits can be enjoyed in months.

In fact, Ryan Schmitz, Manufacturing Execution System Analyst at a large aerospace manufacturer, started his M4.0 journey with a few A&T stations and showed benefits in around six months. Now, 10 years later, he has deployments worldwide.

Incremental steps are good for both learning as you go and for annual budgets. There are specific ways to make this implementation gradually. Also, while the implementation is order independent, it is best to prioritize digitizing the results for early analysis of improvement opportunities.

Recommended Methodology and Building Blocks

But where to start?

An operator at an A&T station typically starts with a work instruction, which lists the steps in the A&T process. Each step creates information, such as when the step was completed, any measurements made, and checks for pass/fail. When the work instruction is completed, the operator creates a record of the results. The desired outcome is to digitize and possibly automate this gathering of information.

Lean initiatives have been in place for decades. Almost every discrete part manufacturer uses Lean at some level.

 

Let’s explore these three A&T activities (work instruction, step execution, recording results) while being mindful of the data consumed and produced to achieve the three M4.0 goals: Lean, digitization, and automation.

1.     Common Format and Metadata

At Viewpoint Systems, experience has taught the importance of flexible yet uniform data descriptions. Hobgood concurs by recommending that any digitization methods and processes have common foundations. For example, all A&T stations could record common items such as:

• The ID of the part about to be assembled and tested
• The name and version of the work instructions
• The timestamps of events (e.g., start of a step)
• Operator and A&T station IDs
• Operating conditions as make sense (e.g., shift number, temperature, power levels)

And so on. This information should be described in a common fashion that enables modification in the future without breaking access to earlier data. XML, JSON, and other text-based descriptions offer enough flexibility without needing to design an entire data management edifice upfront. Data stored in databases should include summary data and pointers to original sources to encourage modularity and maintainability.

An upfront endeavor to identify and accommodate such common data will put you on the right path to applying analytics across your entire factory operations. Schmitz labels these common items as metadata and points out that metadata provides context when analyzing “out of process” events.

2.     Electronic Documentation for Work Instructions

A work instruction organizes the A&T process into a sequence of steps, each describing a small set of tasks. An electronic work instruction is a digitized version of a paper-based instruction manual.

The format of an electronic work instruction can be a PDF, HTML, XML, Word, scanned from existing documents, and so on. An operator using an electronic work instruction follows the same process as with a paper version. Thus, the creation and use of an electronic document, vs. a paper-based version, is an incremental improvement effort.

3.     Automation of Step Execution

The execution of the electronic work instruction can be digitized by using a software application, often called a test executive, to automatically present to the operator each step before moving to the next step. The operator cannot walk through the electronic work instructions in arbitrary order and no steps can be skipped. This disciplined presentation promotes execution uniformity across operators and procedures.

At this stage of the M4.0 journey, any results, such as measurements, verification pass/fail data, comments, and so on, can still be recorded into a paper-based record. The automated, sequential step execution application is an incremental improvement effort that builds on the work done converting work instruction documents to electronic format from paper.

“Common data will allow application of analytics across your entire factory operations.”

 

4.     Electronic Results

The next stage in this endeavor captures digitized data during the electronic work instruction execution. The test executive application is amended to collect step information after presenting each step in the electronic work instructions. A spectrum of automation options exists to digitize these data.

At a minimum, any results previously manually written into a paper-based report will be entered into dialog box field(s) presented to the operator at the conclusion of each step.

At a maximum, the test executive automatically performs the A&T steps. The operator responsibilities are often limited to loading and unloading the unit(s), while full automation has a machine loading and unloading units, a situation usually applicable to very high-volume manufacturing.

In high-mix, low-volume manufacturing, the steps are a mix of these two extremes, with each step being automated at whichever end of the spectrum makes quality and financial sense.

Upon completion of the electronic work instructions, the test executive will corral all the results and send them to an electronic results file, such as a text file, an Excel file, or a database. The formatting and structure of these results should follow the recommendations above about common and flexible data.

Note that the amount and type of information being saved in an electronic record can initially be limited to the details most relevant to your purposes. For example, if you suspect that a subassembly from one of your suppliers is causing issues, capture supplier IDs. Or, if first pass yield is low, gather measurements on unit performance and create a Pareto chart to attack the most prevalent. Schmitz suggests having that product’s manufacturing engineer include a “best guess” at the “out of process” root cause in the unit’s metadata, since it’s harder to diagnosis when only the symptoms (i.e., measurements) are available.

Benefits of A&T Digitization

There are many benefits achievable when implementing the digitization of A&T stations. Remember that common data and methods are required to extract the most benefits from these recommendations.

1.     WIP Visibility and Traceability

Once every A&T station pushes operational data to a central data store, queries could locate a part on the factory floor based on its part ID.

Furthermore, any need to review the manufacturing history of a particular part, including any rework, is satisfied by another query. These queries are often useful for handling warranty issues and when being audited by a regulatory agency.

2.     Identifying and Ranking Sources of Waste

From a Lean perspective, the data collected into the electronic records can be mined and analyzed for operational and the unit performance issues.

Measurements of the A&T execution can be captured, such as the takt time of the entire procedure and each individual step, any failure mode and/or operator comments from each step, the operator involved, and the work instruction used. Statistical analysis of these data boosts the utility of Lean procedures by helping to identify and rank sources of waste. Furthermore, manufacturing managers have access to WIP status and operations managers can identify areas and operators needing improvement and additional training.

“Three M4.0 goals: Lean, digitization, and automation.”

 

For example, analyses of operational data will highlight bottlenecks by operator, A&T station, work instruction step, and by a part type (model). Unit performance outliers and failures can be ranked by step result to find which assembly steps are causing the most problems. Also, any steps that result in failure can be cataloged by failure mode to identify the failures that need most attention.

Without these electronic data sets, choosing where to spend improvement efforts is obviously going to be less efficient than recognizing and attacking the most prevalent issues first.

3.     Training and Skill Management

With a list of training histories for each operator, the test executive can limit access to specific electronic work instructions on which operators have been trained. If operational metrics for one operator show more frequent issues and longer execution times, then there is a training opportunity to improve performance. Or, if the opposite, figure out why this one operator is performing better than others and train the rest.

4.     Version Control

Since electronic work instructions are digital files, version control tools can bring organization and access rights to the versions of those documents used by an operator. Coupling version control with operator access control in the test executive prevents operators not trained on specific product types and models from having access to the wrong versions.

Regulatory audits usually check to assure that the operators know their jobs by checking training records for actual work performed. Controlling access to electronic work records according to the associated training closes this potential audit gap.

The End Game

There are three focus areas for M4.0 efforts on assembly and test stations: Electronic work instructions, automation of assembly and test procedures, and collecting data into electronic results files.

These 3 areas bolster the M4.0 goals of Lean, automation, and digitization by enabling electronic control, generation, and analysis of data.

Each of these areas can be incrementally improved, permitting access to benefits within months of initial efforts, while providing a known path to continual improvement. With an eye towards seeking and deploying commonalities, coupled with flexible data formats, work done on the baseline foundational efforts can be enhanced as time and budgets allow. There is no need to develop a huge monolithic set of tools and procedures.  M

About the author:
James Campbell is Director of Viewpoint Systems.