Gaining the Manufacturing 4.0 advantage with data-driven in-process testing

  • 15-Jan-2018 01:32 EST
Dashboard 3.jpg

An example of a waveform overlay to compare digital process signature data from a production process or test to illustrate the range of acceptable deviation and spot anomalies that may signal a quality issue. In this case, it’s process signatures from a crank torque to turn test for an engine. (all images: Sciemetric)

Manufacturing is changing thanks to the increasingly sophisticated and intelligent use of data to make a production line smarter and more efficient. For off-highway and specialty vehicle manufacturers, the 4.0 revolution offers great opportunity to achieve significant cost reductions and grow revenue.

Whether a company makes machines and equipment for the mining, construction or farming industries, the ability to identify and address defects and quality issues before they make it out to the field is paramount. Most manufacturers Sciemetric Instruments has talked to have a problem process or two they struggle to eliminate that can lead to downstream quality issues. These are problems that hobble their ability to raise yield, reduce costs and increase profitability.

Many Manufacturing 4.0 strategies focus on how to boost productivity and reduce station downtime on the line by monitoring the machines that manufacture parts. But monitoring parts through each step of production or test can also provide crucial, real-time intelligence to get in front of potential quality issues. The digital tools to integrate this connectivity anywhere on the line—to collect, manage and analyze the right data at the right time—are more affordable and effective than ever before.

To start, the team must consider what insight is needed to address a given production issue. Vendors of manufacturing and test equipment are increasingly adding the capability to collect and store data to their machines. While this does provide some value, it results in a build-up of data silos scattered across the plant floor. What a plant needs is data collection and analysis that is integrated into a single database, accessible through a single dashboard, so the team can sort through the noise and find what it needs, when it’s needed most. This is how the true power of data is unlocked.

A true story

Sciemetric worked with one manufacturer of agricultural machinery that struggled to make effective use of its production data without any consistent and centralized means of data management. Scalar pass/fail data from end-of-line engine hot test cells would end up in one silo, entered manually and indexed by time and date stamp. Further up the line, some process stations, such as torqueing for bolts, collected full process signatures, indexed by serial number, but this data ended up trapped in a different silo.

When a product came back from the field due to a customer complaint or warranty issue, it routinely took as long as a week to retrieve all the related scalar and process signature data scattered across the plant.

This lengthy feedback loop created uncertainty and long production delays—the manufacturer didn’t want to risk shipping defective products that would tarnish its brand in the marketplace. In one instance, full production was halted for several weeks.

For an investment that cost only a fraction of a day of lost production, this manufacturer standardized and centralized data collection and reporting across the plant. Quality engineers now had all the data they needed at their fingertips, to catch defects in real-time, trace root cause, and limit the scope of a recall to the specific units the data trail told them were suspect.

Collect the right data

A common starting point to achieve this level of connectivity is to incorporate in-process testing (IPT) into the production line. With IPT, measurement, monitoring, data collection and reporting is built into each step of the manufacturing process. This provides intelligent monitoring and pass/fail determination at the source, as well as acquisition of the data that charts what happened to each part throughout a process.

The result is a consolidated birth history record, indexed by serial number, for every single part. This record includes all the data captured from each process and IPT station on the line that touched a part—captured at the point of its highest fidelity and quality. This data includes scalars, digital process signatures and even machine vision images with their datasets.

It doesn’t matter if that process is a press fit, a weld, a rundown operation, or dispensing a bead of sealant or adhesive—there is a place where a sensor can be added and data collected into a central repository. With this approach, data silos are eliminated.

With today’s intelligent data analysis and visualization tools, the data contained in a birth history record can then be mined at any time for insight. From leaky seals and valves, to bad welds and cracked castings, a problem can be spotted and its root cause determined with data-driven diagnostics and thorough reporting.

This 4.0 insight can be used for much more than catching defects as they occur on the line. It can also spotlight emerging trends and patterns before they become a production issue that leads to higher scrap and rework rates.

A growing number of OEMs, including off-highway, have come to appreciate the value of squeezing all they can out of their technology and their data. To maintain a competitive edge in a Manufacturing 4.0 world, companies need the ability to pluck a needle from the haystack, at any time, on demand, and scrutinize its every detail. IPT and today’s data management and analysis tools offer a relatively easy starting point to boost the intelligence of manufacturing plants.

Dave Mannila, a senior Product Manager at Sciemetric Instruments, wrote this article for Truck & Off-Highway Engineering. He has broad responsibility for new product concept, definition and development, as well as maintaining Sciemetric’s overall product roadmap.

Share
HTML for Linking to Page
Page URL
Grade
Rate It
4.00 Avg. Rating

Related Items

Training / Education
2017-10-19
Training / Education
2013-04-09
Technical Paper / Journal Article
2011-09-13
Training / Education
2017-10-31
Technical Paper / Journal Article
2011-05-17
Training / Education
2013-04-09
Book
2009-03-30
Training / Education
2013-04-09
Training / Education
2017-10-31
Technical Paper / Journal Article
2011-09-13
X