First Article Inspection (FAI) is a manufacturing and quality assurance process for satisfying product specifications and quality requirements before full production begins. It typically involves a detailed examination of the first sample produced from a manufacturing run, known as the “first article.”

This kind of inspection is critical in scenarios where there is:

1. The introduction of a New Product: When a new product is manufactured for the first time.
2. A change in design or process: Whenever there are significant changes to the design or manufacturing process of an existing product.
3. A transfer of production: When production is transferred from one manufacturing facility or production line to another.

The purpose of a first article inspection is to:

• Verify that the manufacturing process can produce components that meet requirements consistently.
• Ensure that any changes are correctly implemented and have the desired effect.
• Identify any issues early in the production process to avoid costly corrections later.

The inspection process includes checking a product’s dimensions, material, appearance and all other relevant features against the design specifications. The process compares the product to drawings, parts lists and other documentation.

Three-dimensional first-article inspections are primarily performed with coordinate measuring technologies, such as stationary, programmable coordinate measuring machines (CMMs), portable CMM arms and non-contact scanners. Two-dimensional measurements are made with hand gages such as micrometers, calipers and height gages.

CMM inspection has many benefits over traditional hand tools, including higher productivity, consistency, accuracy and improved process control. In addition, since 3D CAD/CAM models have effectively replaced blueprint drawings with “digital product definition” (DPD), first article inspection via CMM technology and DPD can be a very productive pairing.

Inspection results are documented in a First Article Inspection Report (FAIR), which is reviewed and approved before proceeding with full-scale production.

Industry Use of First Article Inspection

FAI is crucial in various industrial quality management systems and standards, such as ISO 9001, AS9100 for aerospace, PPAP for automotive industry standards, and ISO 13485 for medical devices. It acts as a form of risk mitigation, as it helps catch potential issues before production begins.

The term “first article inspection” applies to a standard operating procedure used in the aviation, space, and defense industries, and is defined by SAE specification AS9102 Aerospace First Article Inspection Requirement, written and controlled by the International Aerospace Quality Group (IAQG).

For the automotive industry, the Production Part Approval Process (PPAP), defined in detail by the Automotive Industry Action Group (AIAG), is the well-established practice of inspecting products for design and manufacturing verification in the early stages of production. First article inspection is part of PPAP or may precede full-blown PPAP. The difference is that FAI is conducted on the first part or parts produced from a new or changed design, while PPAP is a more extensive process evaluating a series of parts from a full production run for new parts.

The medical products industry requires manufacturers to put FAI or something comparable to FAI into practice by requiring a process that verifies product compliance with the ISO 13485 Medical Devices – Quality Management Systems specification.

The Aerospace First Article Inspection Requirement –AS9102

First article inspection (FAI) per AS9102 is an established process and requirement for aviation, space, and defense manufacturing sub-contracts. Having been proven and refined over many years, it has also become a reliable best practice for other industries.

FAI is enforced as a contractual requirement dictated by strict conformance to the SAE AS9102 specification, mandated to ensure fully verified and documented quality as a new product begins manufacturing.

(Image: Verisurf.)

Safety of flight issues present the primary need for such diligence and adherence to a controlled process. Furthermore, consistency across a widely distributed supply chain with methodology that is scalable and reliable must be proven to be achievable by all stakeholders.

The AS9102 details the full scope of activity and documentation required, even establishing a set of forms that ensure all aspects of design, DPD, manufacturing, materials, engineered processes, testing and inspection are verified and recorded so that nothing gets overlooked.

Dimensional inspection typically requires the most time and effort of all the FAI tasks. While digital definition is an advancement for improving productivity throughout the product lifecycle, it has created the need for additional controls to be added into the FAI process to ensure accuracy and consistency as well as conformity across the supply chain. The “digital thread” must stay intact, and controls are required to prevent errors, especially when handing off digital information from one platform to the next. Appropriately, the AS9102 specification has been updated to include the necessary controls.

In more recent times, the aerospace industry has begun to introduce its supplier base to automotive industry-style PPAP requirements. This can be seen in aerospace standard AS/EN/JISQ 9145:2016 from the IAQG and confirmed by Boeing X38656 for Advanced Product Quality Planning (APQP) and Production Part Approval Process (PPAP).

The Automotive Production Part Approval Process (PPAP)

The purpose of PPAP is to determine if all customer engineering design and specification requirements are properly understood by the supplier and whether the process yields products consistently meeting the requirements during an actual production run at the quoted production rate. It validates more than just the part, but also the entire engineering, manufacturing, supply chain and quality control aspects of the product lifecycle.

Like FAI, PPAP requires full dimensional inspection and reporting, among other requirements.

FAI-type inspection is a part of the PPAP but differs from aerospace due to the unique nature of automotive manufacturing, produced at much higher volumes, differing government regulations, aesthetics of part surfaces, legacy industry methodology and expectations. With eighteen elements identified in the PPAP framework, the quality control plan, measurement system analysis and dimensional results are among the drivers that lead manufacturers to a similar dimensional inspection process.

(Image: Verisurf.)

Medical Quality Management Systems – ISO 13485 for Medical Devices

First article inspection is a necessary step to ensure a manufacturing process will create a medical device that works as designed and meets product requirements. The data and measurements gathered from FAI become part of the medical device file, which is required for FDA submissions and ISO 13485 compliance.

Even though ISO 13485 is a stand-alone standard, it is based on ISO 9001. So, while ISO 9001 is an internationally recognized standard for any organization in any industry, the ISO 13485 standard includes added requirements that are specific for companies that manufacture ISO-compliant medical devices.

Planning for FAI

As part of an organization’s quality management system (QMS), manufacturers adhering to ISO standards need an overall quality plan to monitor and control all aspects of part production and an inspection plan that focuses on measurement throughout the product lifecycle. The AS9102 specification requires manufacturers to have an FAI plan that is specific to new production runs.

As a sub-contract requirement, source inspectors want to see that their suppliers are not just checking parts, but have a documented process for inspection, management of non-conformances, and corrective actions when necessary. The inspection plan is based on the technical data package tied to the customer contract.

The manufacturer’s QC department uses drawings, material and process specifications, bill of materials (BOM), digital datasets such as CAD, and other sources from which to create the inspection plan. The manufacturer must follow and document all DPD protocols (for example, ASME Y14.41 & Boeing D6-51991 specifications) to ensure there is no disconnect within the product definition exchange to prevent incorrect or superseded data from being used erroneously.

The manufacturer’s QMS establishes the policies and workflows for their inspectors to ensure compliance with the appropriate standards. Compliance auditors will verify that the quality manual and FAI procedures are followed and meet the applicable requirements.

(Image: Verisurf.)

First Article Inspection Report (FAIR)

To make sure that all important characteristics are verified, there must be no disputing what the baseline (nominals) and tolerances for those characteristics are. What is acceptable and what is not must be clear. Therefore, thorough and traceable documentation is necessary and is the main purpose of the first article inspection report (FAIR). It is both the sub-contract deliverable as well as the historical record, which can serve a number of purposes down the road.

Included in a complete FAI is an accounting of all manufacturing processes and equipment involved in manufacturing, planning, procurement, sub-contracting and quality assurance of the product. It requires confirmation of equipment maintenance (such as machine calibration and certification), bill of materials, chemistry, heat treatment, surface finish, material certifications, material properties and purchased parts.

(Image: Verisurf.)

The AS9102 specification establishes a standardized format ensuring that the FAI process is complete, and that the foundation is set for apples-to-apples comparisons between successive parts.

The AS9102 standard requires three forms to be included in the FAIR:

• Form 1 – Part Number Accountability
• Form 2 – Product Accountability – Materials, Special Processes and Functional Testing
• Form 3 – Characteristic Accountability, Verification and Compatibility

With the primary focus of this article being dimensional inspection; we will concentrate on Form 3 in what follows.

Form 3: Characteristic Accountability Verification and Compatibility Evaluation

The AS 9102 Form 3 is the record of characteristics measurements alongside associated design values and tolerances in a line-by-line chart form. Each record also notes the characteristic’s either conforming or non-conforming condition. For the most part, all drawing dimensions and specified requirements are included. Each characteristic gets its own line labeled with an ID number (some line items report on a group of features when necessary). Line items from Form 3 coincide with characteristics noted on the engineering drawing or DPD with a circled “balloon” ID number. This is from where the term “ballooned drawing” or “bubble drawing” often referred to by those involved comes from.

In the ballooning process, inspectors create the sequential encircled numbers on the drawing next to each dimension or GD&T callout. The importance of Form 3 should not be underestimated, which is why each measurement or observation should be clear and unmistakable. Each design characteristic needs to be easily understood, tolerances or allowable conditions should be clearly noted, non-conformances should stand out on the page and the values showing the degree of in or out of compliance must also be shown

When the available product engineering is a completely CAD-model-based, digital definition, ID balloons may reside within the model where they can be associated with features. GD&T that is connected to the entity for which it controls, and is usable by the measurement software, is referred to as “associative” or “semantic.”

Measurement software such as Verisurf offers associative/semantic GD&T, thereby preserving the digital thread. (See Verisurf Model-Based Definition.)

Form 3 fields and column headers are differentiated by typeface style (bold, bold italic, or standard font) as either required, conditionally required or optional:

• (R) – Required: This is mandatory information. NOTE: These fields are depicted in bold font.
• (CR) Conditionally Required – This field shall be completed when applicable to the product (e.g., serial number shall be entered when there is a serial number) or when required by the customer. NOTE: These fields are depicted in bold italic font.
• (O) Optional – This field is provided for convenience; the field may be left blank. NOTE: These fields are depicted in standard font.

Form 3 – Characteristic Accountability Verification and Compatibility Evaluation Ballooned Drawing. Verisurf software integrated FAIR features – Verisurf metrology software automatically creates balloon IDs and assigns them to dimensions and callouts in the 3D model. (Image: Verisurf.)

The Dimensional Inspection

First article Form 3 inspection mostly involves dimensional measuring. (There may also be some design notes and quality process notations that need to be observed and reported as well.) The engineering drawing or DPD model dimensions have an associated tolerance connected to each dimension or GD&T callout/control frame.

In some cases, tolerances are controlled by the number of decimal places shown on dimensions which in turn correspond with the tolerances noted in the drawing title block key for each number of decimal places. In effect, the FAI requires that every dimension and callout be measured, recorded and compared to the nominals (as-designed values). GD&T control frames, per ANSI specification Y14.5, dictate the feature relationship constraints that are to be measured and reported.

For this important step of the FAI, the inspector measures, calculates per GD&T guidelines and reports each value on Form 3. On each row of the spreadsheet, design values and tolerances are recorded in column 8 “Requirement,” measured values are recorded in column 9 “Results,” and “Pass” or “Fail” noted when applicable.

(In some cases that do not involve measured values, reference to the statement of conformance and indication of pass/fail or “accept” may be all that is recorded.)  

If there are any that fail, a non-conformance number is assigned in column 11. The FAI would be identified as “not complete” and so noted on Form 1 until the issues are corrected and the FAI or partial FAI is completed with no non-conformances.

Generally, measurements are to be recorded as individual values for each characteristic. However, if a group of measurements is identified in the requirements, such as callouts for surface profile or hole position, the minimum and maximum may be reported as long as all measurements are within tolerance. If any are out of compliance, the measurements that are non-conforming need to be singled out individually on their own line.

For positional tolerance, CMM measured nominal and measured point values are not sufficient, and the actual positional nominal, tolerance and deviation need to be reported.

(Image: Verisurf.)

Automating First Article Inspection Ballooned Drawing Reporting

FAI measurement, recording, and reporting by hand is a tedious process, but several first article inspection software offerings, such as SOLIDWORKS Inspection, have been developed to automate the process.

These products use optical character recognition (OCR) to read drawing dimensions and GD&T, and automatically add sequenced balloon IDs to each. This data is formatted to a spreadsheet which can be printed, output to a shareable file, or fed into a database for SPC and subcontract management. The complete digital process reduces transcription errors and improves productivity while supporting the digital thread.

While inspection using hand tools and manual gages still plays a big role in FAI, modern CMM technology helps to achieve accurate, digitally recorded measurements often in less time than it might take just for the setup for a complex GD&T measurement using manual methods. Inspection with a CNC or DCC, motor-driven CMM is automated by simple, intuitive selection of features and path editing in the CMM software.

Productivity gains can also be achieved by using manual CMM devices such as portable articulating arms controlled by automated, software-prompted program plans. The efficient, digital workflow modernized by applications such as Verisurf software, provides automatic program path creation that can streamline the process even further.

Regardless, automation through either motor-driven CMMs or software-prompted manual CMMs provides productivity enhancements and improved process control over ad hoc, old-school, manual gage measuring.

(Image: Verisurf.)

Gaining Source Inspector Buy-Off

A major goal of all suppliers is to get source inspectors to buy off first articles so their organization can get on with production and make a profit. They want source inspector visits to go as smoothly as possible to strengthen the company’s reputation and for opportunities to get more contracts. Automating the FAI and CMM inspection helps in this regard by demonstrating a commitment to modern quality processes.

The supplier’s FAI process itself needs to be well-documented and controlled to ensure that it comes off sparkling clean and professional. Having a strong checklist, diligent preparation before the source inspector’s visit and paying attention to fine details are key. Doing so will make it routine, and the feedback you get from the process will help to make your systems more profitable. To improve your preparation, FAI automation software along with high-performing CMM software will smooth your source buy-off.

Embrace First Article Inspection

Some may see mandated first article inspection as burdensome and costly, so they unenthusiastically comply as part of their sub-contract requirements. If we keep our eyes fixed on the benefits of FAI being excellence in process control, reduction in scrap and superior customer satisfaction, we will see it as being completely worthwhile.

Recognizing the peace of mind of knowing we contributed to the safety of our customers and the public, while ensuring there are minimized chances for supplier-customer disputes down the road, only adds to our comfort level.

Here is a summary of first article inspection tips that should be of help:

• Ensure all of the FAI requirements are clearly defined in the contract.
• Insist on specific and well-documented details of all the engineering characteristics that will need to be verified in the FAI report.
• Seek clarification for all digital product definition data and traceability requirements and make sure they are documented. Verify that there are no conflicts between DPD and physical engineering/contract definition and which is the contractual controlling authority.
• Prepare a comprehensive FAI plan before commencing. Make sure the applicable FAI specification (such as the AS9102) is in hand and compliance is an unmistakable part of the plan. It should be documented in your quality management system handbook.
• When any questions arise about what makes an acceptable product, or if manufacturing challenges arise during production, communicate with the customer and get them resolved (in writing) at the earliest opportunity. Make sure to convey to your organization that everyone is on the same team with the same goals in mind.
• Prepare for source inspection (if applicable) with reckless abandon like your life depends on it. You don’t want any friction with the source inspector, and a compliant, respectful collaboration is essential. If your in-house FAI source inspection representative isn’t a “people person,” hand the job off to someone else.

Learn more at Verisurf.

The post First Article Inspection with 3D Measurement Systems appeared first on Engineering.com.

Software drives workflows and ensures repeatable process control. Select your software first, then decide the type and brand of CMM you need for measurement and data collection. (Image: Verisurf.)

There are best practices for selecting software, and that includes measurement and inspection solutions for manufacturing enterprises. Too often, manufacturing companies either minimize the importance of measurement and inspection software or, worse, they make purchase decisions based on hardware features that may or may not provide the results they need.

Only after taking delivery and using the equipment do they realize they have purchased software that is not up to the task, or the wrong type of coordinate measuring machine (CMM) for their workflow and the results they require.

The goal of this article is to help you avoid making the same mistake, through a comprehensive approach to software selection with critical consideration given to needs assessment, compatibility and extensibility. If you have a clear perspective of your needs and a better understanding of what questions to ask and what features to insist upon, the result will be a grounded solution that provides results, enhanced workflow and added value across your manufacturing enterprise.

Assessing Needs and Communicating

You have probably heard it said that if you don’t know where you’re going, any road will take you there. This holds true for selecting the right measurement and inspection software for your business, as well.  Before purchasing any software, regardless of application, you need a clear assessment: 

• What problem or opportunity are you addressing with technology?
• How are you getting the data you need, today?
• What is your workflow?
• What benefits will be expected from your new software?
• Most importantly: Who else in the organization will be affected?

When it comes to measurement and inspection, there are a few considerations that are overlooked surprisingly often. These include identifying the maximum part volume to be measured, understanding the number of measurement points needed to calculate results and the speed required to complete the measurement process.  Failure to consider these points may lead to a solution that is inefficient or, worse, one that does not support a critical workflow.

Here’s what you don’t want to do: see an impressive CMM and buy it based on a demo, possibly unrelated to your workflow or measurement objectives, then try to justify why it is the best solution for your organization.

That’s putting the cart before the horse.

This is why clear communications should start with your assessment and carry through the balance of your technology acquisition and implementation.  You might be surprised how many people in your organization can be positively impacted by data gleaned from measurement and inspection software. 

Investing in the right set of tools can mean using that data for quality inspections, reverse engineering, recreating data for missing features, support for tool making, and guided assembly of components, fixtures and weldments. Collected data can feed downstream applications, such as Product Lifecycle Management (PLM) or Statistical Process Control (SPC) databases.  Understanding who the stakeholders are and keeping them in the loop can add a great deal of value to your software. 

When selecting measurement and inspection software, make sure that it not only is compatible with CAD but is CAD-based at its core. A good question to ask is, “Does this software include intelligent 3D CAD modeling capabilities?” (Image: Verisurf.)

Tip #1 – Make a CAD Connection

In today’s business environment, every software platform must play well in the sandbox with already-established enterprise software or protocols. For business intelligence, this may be enterprise resource planning (ERP) software. For business communications, it might be Microsoft. Within the manufacturing enterprise, the standard is CAD. 

Finished products and components are typically conceived in a CAD environment and, for most manufacturers, they remain in this virtual state throughout the design-build process.  Data continues to be added to the file and its digital references throughout the lifespan of the product.

Model-Based Definition (MBD) is the term given to this commitment to an intelligent CAD model. The central concept embodied in MBD is that the 3D model provides all the detailed product information necessary for all aspects of the product life cycle. In the example of a machined component, this would include intelligent Geometric Dimensioning and Tolerancing (GD&T) annotations.

This removes the ambiguity, conflict and doubt that arises when drawings and models co-exist. By bestowing authority on the model, MBD eliminates errors that result from referencing an incorrect source and makes processes more efficient—no more searching to determine correct revision levels.

When selecting measurement and inspection software, make sure that it not only is compatible with CAD but is CAD-based at its core.  A good question to ask is, “Does this software include intelligent 3D CAD modeling capabilities?”

A truly CAD-based platform is the best way to ensure data continuity (digital thread) and maximize model-based measurement and inspection applications across your manufacturing enterprise.  Even if your organization is not at the point of fully embracing MBD, you want software extensibility and room to grow.

Tip #2 – Ensure Software Compatibility

If you make CAD-based measurement and inspection a requirement and do your homework, you’re already halfway covered with respect to compatibility.  The CAD connection will allow you to import, edit, save and export to and from virtually any CAD software. 

This is important when it comes to creating inspection plans that might have missing GD&T data that needs to be added, or when using reverse engineering to fill in missing features. 

Having an open CAD platform from which to operate and manage your measurement and inspection applications makes downstream integration and data sharing much easier with common file types and established transfer protocols. 

When evaluating measurement and inspection software, universal compatibility with all CAD software and all new and legacy hardware is essential for data continuity, repeatable process control and optimum resource utilization. (Image: Verisurf.)

Tip #3 – Don’t Forget About Hardware Compatibility

There are many types and brands of CMMs available on the market and if you remove the software from the equation, they all do one thing very well: collecting precision measurement points. 

The power behind all CMMs—contact and non-contact, fixed and CNC-type, portable arms, trackers and more—is the applications and workflow provided by the supporting software.  Most CMMs come with some sort of measurement and inspection software, but these are by CMM manufacturers, not software developers.

CMM manufacturers tend to be product-centric when it comes to software, focusing only on their machines.  It is for this reason that you need to keep your evaluation of software and hardware separate and stay true to your assessment plan. 

Consider your workflow and corresponding software applications first. This will lead you to the correct type of CMM you need before deciding on a brand of CMM.

When looking at measurement and inspection software from a hardware compatibility standpoint, it is important that the software can drive all CMMs and import measurement points in real-time.  This is especially significant when collecting and processing very large measurement sets encompassing millions of measurement points, typically associated with non-contact scanning of large complex surface profiles.  The software must be able to keep up with that influx of data or your productivity will grind to a crawl.

Universal device compatibility is critical when it comes to measurement and inspection solutions for a manufacturing enterprise. In the case of inspection, workflow routines should be able to be programmed once and then executed using any available CMM. This provides flexibility and consistency among operators and ensures repeatable process control is maintained, which is a key tenant of quality management.   

Tip #4 – Ask About SDKs for Integration

Selecting a measurement and inspection software that is based on an open platform with built-in compatibility and interoperability is critical to advancing the concept of the digital thread. 

To be effective, each step of the design/build process must play its part to stay connected.  Data in, data out, and the ability to pass information across all manufacturing applications and databases are important. A good question that will confirm a software’s commitment to interoperability is, “Do you offer a Software Development Kit (SDK).”

If the answer is no, integration might be harder than you think. 

Tip #5 – Remember Training and Maintenance

Training and compatibility go together. 

Over the last 10 years, there has been a surge of shops adding portable measurement and inspection assets.  These newer devices are based on open software compatibility, but many of the legacy CMMs that are still in use are based on closed systems and require dedicated software to run them. 

This has created a dichotomy in some shops between new and legacy systems.  Certain quality inspectors are trained to operate legacy machines and others are versed in newer equipment.  This causes bottlenecks due to inefficient resource utilization.

Your software selection should be enterprise-based: one software to drive all your CMMs (regardless of type or age), manage workflow and provide your desired application results. This will help with resource utilization, reduce bottlenecks and lower software training and maintenance costs, since you don’t have to support more than one platform. 

Bonus Tip – Ask for Help

When it comes to selecting inspection and measurement software, getting it right is important, so don’t be hesitant to ask for help.

 Software and hardware providers alike should be willing to take an interest in your business and help you put together the right recommendation for your unique application.  Just try to remain objective and stick to your goals. 

There are many measurement and inspection solutions available, so hold out for the one that fits your needs.

Learn more at Verisurf Software.

About the Author

Nick Merrell is Executive Vice President of Verisurf Software, Inc.

The post 5 Tips for Sourcing Measurement and Inspection Software appeared first on Engineering.com.

Vehicle inspections are crucial for both quality control and maintenance, but they take time. And in the bustling environments of manufacturing or fleet management, every second counts. Finding ways to reduce inspection times without compromising standards or (worse) overlooking a serious defect is a difficult balancing act that companies spend hundreds of millions of dollars trying to get right.

New Jersey-based UVeye has spent almost a decade developing its own unique approach to vehicle inspection using computer vision. With funding from Toyota Tsusho, Volvo and Hyundai Motors, the company has developed tools for inspecting vehicle undercarriages, outer bodies and interiors. Now, Amazon joins UVeye’s growing list of customers, which also includes CarMax and GM.

According to a press release from UVeye, the startup will begin rolling out its vehicle inspection systems across hundreds of Amazon warehouses in the U.S., Canada, Germany and the U.K. This follows a trial phase, during which the UVeye system identified issues that are often overlooked during manual inspections, including damage to the undercarriage and nails in the tire treads.

“Each new feature and improvement has brought us closer to where we are today, and we are so proud of the system, which is now being scaled to administer tens of millions of inspections a year,” said UVeye co-founder and CEO Amir Hever in the press release. UVeye claims that its technology is currently used to scan over 2 million vehicles and 20 million tires every year.

Tom Chempananical, Global Fleet Director at Amazon Logistics, echoed Hever’s sentiment in the same release: “With over 100,000 Amazon-branded delivery vehicles on the road, bringing 20 million packages to our customers every day, we can automate most of the inspection process at scale,” he said. “This reduces the time spent on inspections by DSPs and delivery associates, ensuring packages reach customers faster while improving road safety.”

According to Amazon, a typical manual inspection takes approximately five minutes, while the UVeye system takes less than a minute to complete the inspection process. The ability to shave those valuable minutes from each inspection makes this partnership between the two companies a natural one, especially since the data that UVeye’s system collects and the AI software that processes it both run on Amazon Web Services.

The post Amazon to Deploy Automated Inspections for 100,000 Delivery Vehicles appeared first on Engineering.com.

VXinspect enables you to validate the quality of manufactured parts and generate reports based on thorough engineering requirements. (Image: Creaform.)

As the manufacturing world pushes towards the goal of zero-defect production, part inspection is critical. Sheet metal’s thinness, coupled with its susceptibility to warping, demands precision inspection—even the most minor deviations have significant implications down the line. This underscores the need for exhaustive and precise quality control reports.

Sheet metal components have certainly evolved along with other technological advances in manufacturing to handle complex and intricate parts that traditional inspection methods now struggle to measure. The pursuit of zero-defect perfection has driven the development of quality control methods with cutting-edge technology that makes use of the vast amount of available data to navigate complex designs, quickly becoming the standard in the manufacturing world.

The post How to Improve Sheet Metal Inspection appeared first on Engineering.com.

This summer, the legendary Danish toy manufacturer LEGO announced a breakthrough: lasting bricks made from recycled PET, derived from soda bottles. The research project was considerable, requiring a staff of 150 working through 250 different plastic compositions to create a recycled plastic brick that met LEGO’s tough quality standards. Although an engineering success, on September 25, the company announced the suspension of the project, as an analysis of the carbon footprint of the recycling process showed no net benefit compared to oil-derived plastic resin. The company is switching research efforts to versions of resin derived from e-methanol, feedstock made from waste CO2 and hydrogen. Made from hydrogen derived from electrolysis powered by clean energy, this could prove to be truly zero carbon polymer resin.  

Access all episodes of This Week in Engineering on engineering.com TV along with all of our other series.

* * * 

Episode Transcript:

For decades, plastics engineers working in the injection moulding industry acknowledged that one of the global leaders in advanced technology and efficient production is a very familiar Danish company: LEGO.  

The iconic toy resin bricks have been a staple under Christmas trees for decades, and the relative handful of simple shapes that emulated building bricks have been joined by a stunning array of complex, moving parts. Lego is as much a culture as a toy, and like many European manufacturing concerns, the company is addressing environmental issues. A net zero carbon footprint is promised by 2050, and the company has a realistic action plan to achieve this goal. 

Another angle toward sustainability has been the company’s experiments in using recycled resins for its plastic bricks. On the 23^rd of June, LEGO developed a brick made from recycled PET resin, derived from soda bottles. The research effort was considerable, involving a team of 150, and testing over 250 different plastic resin compositions.  

The recycled resin used was sourced from U.S. suppliers and was approved by the FDA and the European Food safety Authority. A single 1L PET soda bottle contained material for about 10 small Lego bricks.  

But in a surprise move announced on September 25, Lego is suspending the recycled plastic initiative. From an engineering perspective, the project was a success, producing plastic parts with the necessary durability and physical properties to match LEGO’s standards. Although an engineering success, the overall carbon footprint of recycling the PET bottles, then reprocessing them into clean, injection-mould-ready regular resin, was no better than that of the current petroleum-derived version thermoplastic.  

The company announced that it still intends to make Lego bricks from sustainable materials by 2032. The LEGO experiment was an engineering success, but failed to meet its objective of lowering carbon emissions.  

Well, that’s a failure in one sense. It represents a sea change in the way environmental issues are handled in the manufacturing of mass production consumer goods. The carbon footprint of recycled products must represent the entire production process, and when selection, shipping, sorting, cleaning and reprocessing are factored into the equation, it’s possible that recycling is not only not environmentally more sustainable, but may be a net negative if the oil-derived version resin is recycled at the end of its lifetime, and repurposed into products where the carbon balance is net negative.  

LEGO is an especially demanding user of plastic resin, operating some of the world’s most advanced injection moulding equipment to produce low price consumer goods with consistently outstanding surface finish and dimensional stability. For firms that don’t require the quality attributes necessary at LEGO, recycled resin may still provide a carbon benefit.  

In the meantime, LEGO is investigating thermoplastics derived from e-methanol instead of crude oil. E-methanol is generated by reacting waste carbon dioxide with hydrogen, which itself can be generated through electrolysis, powered by green energy. These e-resins are definitely carbon neutral, but cost remains a concern.  

LEGO is a toymaker, but in the plastic injection moulding sector, when they talk, people listen. 

The post LEGO Abandons Recycled Plastic Project. Why? appeared first on Engineering.com.

3D computed tomography (CT) is essential in nondestructive testing for quality engineers in many industries. Creating a 3D image of a sample by compiling multiple X-rays from different directions around a rotational axis results in highly detailed representations of both internal and external features. This can reveal otherwise hidden flaws, and is also useful for geometric dimensioning and tolerancing analysis.

However, one of the drawbacks of using CT is that the images it produces can have limited resolution because the axis of rotation must be orthogonal to the x-ray beam. As a result, images of samples with an area of interest at their center may lack the necessary level of magnification to improve resolution because moving the sample closer would cause it to collide with the x-ray source as it turns.

Computed laminography is a subtype of x-ray tomography in which the rotation axis of a sample is titled at an oblique angle to the x-ray beam. It’s sometimes referred to as “2.5D inspection” because it lies between 2D x-ray radioscopy on the one hand and CT on the other. While it’s most commonly applied to flat components, such as printed circuit boards (PCBs) and microchips, CL is useful for any component with a high aspect ratio, with long x-ray path lengths along the object plane or where the size of the component requires the rotation axis to be positioned further from the x-ray source to avoid collision.

Tilted CL gets around this issue with CT by allowing the axis of rotation to be adjusted up to 30 degrees, enabling the sample to rotate fully. This results in both higher magnification and image clarity, as well as faster scanning times. A company press release claims that, “in one comparative test, a scan that took more than seven hours using an X-ray microscope was completed in less than one hour with Tilted CT on a SEMI S2/S8 compliant XT H225 2x system.”

The technology can also eliminate artifacts caused by high-density features that lower x-ray attenuation. Tilting the axis of rotation helps position areas of high-attenuation so that they rotate above or below a lower-density area of interest, rather than in front of it.

As a result, Tilted CT can aid the inspection of additively manufactured metal parts before they are separated from the build plate. Tilting the entire structure, including the plate, ensures that the denser plate is rotating at the same angle as the x-ray cone beam, thereby eliminating it as an obstruction.

In addition to the XT H225 2x system, Tilted CT is also available for Nikon’s XT H225 and large-envelope M2 X-ray CT systems. While it’s technically applicable to any size of component, users are still restricted by cabinet size and the requirements for holding a stationary sample at an oblique angle. As such, samples more than half a meter in length will likely be difficult to hold still during scanning.

Watch the introduction video below for more information:

The post Nikon’s Tilted CT Takes a New Angle on Nondestructive Testing appeared first on Engineering.com.

Although there is no universal certification body, there are dozens of organizations that offer training and accreditation for Six Sigma experts, including the American Society for Quality, The Chartered Quality Institute and more than fifty universities worldwide. Then there are the consultants and advisory services, from independent specialists on Upwork to industry titans like Bain and Company.

The benefits of Six Sigma tend to be framed in startling figures: billions of dollars in cost reductions supposedly achieved through double-digit gains in production efficiency. And yet, for a management philosophy so focused on statistics, independently verifiable evidence for Six Sigma’s effectiveness seems sorely lacking. What we have instead is mostly anecdotal and, as it’s been observed: the plural of anecdote is not data.

Given the rise of Industry 4.0 and resulting changes in the landscape of manufacturing, it’s worth asking whether Six Sigma’s principles and practices are still relevant today.

Six Sigma in 100 Words

Six Sigma refers to the fraction of a normal distribution curve that lies within six standard deviations of the mean. This represents the target defect rate for a manufacturing process, which works out to 3.4 defects per million opportunities (DPMO). This target is achieved through a combination of various tools and methodologies, such as DMAIC (Define, Measure, Analyze, Improve and Control) for existing processes and DMADV (Define, Measure, Analyze, Design, Verify) for new ones. There are also tools that can be found in other quality management systems besides Six Sigma: the 5 Whys, fishbone diagrams, control charts and cost-benefit analyses.

Who Uses Six Sigma?

Although it began at Motorola, Six Sigma quickly spread to many other large American companies, including Johnson & Johnson, Texas Instruments and, perhaps most famously, GE under the leadership of Jack Welch. Welch is often credited with spreading enthusiasm for Six Sigma after implementing it in GE’s manufacturing operations in 1995.

While the list of companies that claim successful Six Sigma implementations seems impressive at first blush, it’s worth noting that industry winners like Amazon and Google are balanced out by losers like Sears and Computer Sciences Corporation. This raises the question of whether Six Sigma advocates are guilty of a kind of survivorship bias. After all, even the originator of Six Sigma—Motorola—still ended up with staggering losses that led to the company splitting.

Nevertheless, Six Sigma companies do share certain features in common. Chief among these is being relatively large in size, with even the most diehard Six Sigma fans admitting that the philosophy is best applied by organizations with more than 500 employees. Of course, they’ll also add a caveat that there are still many Six Sigma tools and techniques that can work for small- and medium-sized enterprises (SMEs). Why waste a chance to recruit more acolytes to The Six Sigma Way?

Criticisms of Six Sigma

While there have been numerous complaints about Six Sigma over the past few decades, they can be roughly grouped into three broad categories: Applications, Effectiveness and Innovation. In each case, these critiques boil down to questioning whether Six Sigma is really worth all the trouble involved in implementing it, though they take different perspectives on where the trouble lies.

Applications of Six Sigma

In a book published the year after Jack Welch deployed Six Sigma at GE, quality consultant Philip B. Crosby pointed out that Six Sigma’s 3.4 DPMO standard was woefully inadequate for semiconductor manufacturing. Since semiconductors contain more than a million components, each of which needs to be defect-free for the entire unit to function, 3.4 DPMO simply won’t cut it. That was in 1996, and semiconductors have only gotten more intricate and complex in the decades since.

Many other complex products have entered the market since Six Sigma was first conceived, including those which use semiconductors as essential components. Given the growing complexities of manufacturing, it’s worth considering whether a quality management philosophy that predates the fall of the Soviet Union is still applicable in the age of Industry 4.0.

Six Sigma’s Effectiveness

Given all the enthusiasm for Six Sigma, one might expect that there would be plenty of evidence for its effectiveness as a quality management strategy. As noted above, there are assertions that companies which have deployed Six Sigma have seen hundreds of millions or even billions of dollars in cost savings as a result—but there are several reasons to be skeptical of such claims.

Satya S. Chakravorty, a professor of operations management at Kennesaw State University, and Praveen Gupta, director of quality for the Stephen Gould Corporation, have separately claimed that over 60 percent of corporate Six Sigma initiatives fail. According to Chakravorty, this is typically due to external Six Sigma experts who initially spearhead these projects but leave before they’re completed. Charles Holland (who, it should be noted, advocates for a competing quality management system) has argued that 91 percent of large companies that announced Six Sigma programs have trailed the S&P 500 since doing so.

The trouble is that the success (or indeed criticisms like Holland’s) of Six Sigma projects may suffer from the old informal fallacy of post hoc ergo propter hoc: after this, therefore because of this. As Yasar Jarrar and Andy Neely put it in their 2003 critique: “[A]re we making a true improvement with Six Sigma methods, or just getting skilled at telling stories?”

Does Six Sigma Stifle Innovation?

Another of Holland’s (admittedly biased) criticisms of Six Sigma is that it was narrowly designed to fix existing processes and, as such, allows little room for new ideas. Jarrar and Neely make similar claims, noting that most organizations are neither designed nor led to allow the kind of statistics-driven management that Six Sigma requires. As a result, incorporating Six Sigma into a company’s culture may inhibit innovation, either because its tools and methods are too rigidly applied due to a lack of internal expertise, or because the bureaucracy of Six Sigma forces all projects to conform to its particular format.

Chakravorty’s research has found that Six Sigma programs can end up adding more work—in the form of statistical analysis—to teams that are supposed to be eliminating waste, not squeezing more tasks into their daily routines. To be fair, this is more of a risk for companies that try to use Six Sigma by enlisting the temporary aid of consultants, rather than relying on internal experts or champions. Nevertheless, given the abundance of companies and specialists offering Six Sigma services, the likelihood of this outcome is certainly nonnegligible.

Is Six Sigma Worth It?

Although it’s obviously different from purchasing a new piece of equipment, such as a coordinate measuring machine (CMM) or a laser scanner, implementing Six Sigma does require a significant capital investment. Whether contracting with external advisors or consultants or building the expertise internally by hiring candidates with Six Sigma certifications or paying for training, Six Sigma doesn’t come cheap.

Is it likely to deliver a sufficient return on investment to justify that upfront cost?

Ultimately, the answer to that question depends on what you’re trying to achieve with Six Sigma and what kind of guarantees its practitioners can offer that isn’t laced in jargon and carefully curated statistics. As Terry Pratchett once wrote: “Always be wary of any helpful item that weighs less than its operating manual.”

The post Is Six Sigma Still Relevant in Modern Manufacturing? appeared first on Engineering.com.

IMAGE: Sandia National Laboratories

They may be in less demand today than during the COVID-19 pandemic but N95 respirators continue to be a crucial piece of personal protective equipment (PPE). For millions of people worldwide – from healthcare and construction workers to the immunocompromised and their families —having access to high-quality masks is simply nonnegotiable.

Given those stakes, N95 mask tests must be not only reliable but also quick enough to keep up with demand – which isn’t the case with existing systems. But this month, a team of engineers and scientists at Sandia National Laboratories revealed a promising new approach. Combining creative design with 3D printing, it could be the foundation of an all-in-one testing system that produces safer masks, faster.

Problems with the Current Approach to N95 Mask Testing

The core of an N95 respirator is a fine mesh of nonwoven polypropylene fabric produced by melt blowing, which forms an inner layer to filter out hazardous particles. Traditional testing of this core is “very time-consuming and not as efficient as it could be,” Michael Omana, an aerosol scientist at Sandia National Laboratories in Albuquerque, New Mexico, said in a press release. It involves using hot wax or putty to attach a mask to a flat plate enclosed in a box, introducing a test aerosol, and then measuring penetration levels. Because National Institute for Occupational Safety and Health requires 20 masks in each batch to be tested before the batch can be approved for use, the testing phase represents major bottleneck in the supply chain.

Beyond being time-consuming, this approach only tests the mask’s inner layer. “It doesn’t test geometry, how the respirator fits on a face, how it’s moved on and off multiple times, how the straps perform, how the nose bridge performs, how the mask can wear over time,” said Todd Barrick, one of the engineers on the research team.  Added Omana: “I think there were a lot of lessons learned with everyone suddenly looking at what the industry standards were.”

Testing N95 Respirators Holistically

To begin, the Sandia team replaced the flat plate with a 3D-printed model of a human face that can be loaded into a commercial filter test system. According to engineer and team member Brad Salzbrenner, the idea is to test more performance factors at once, such as how the mask fits. Using 3D printing enabled the team to make the model pliable to more closely resemble human skin.

This was a good start, but it didn’t fully account for real-world usage, so the team developed a more complete model of a human head that can be placed in an airtight box and then loaded into the testing machine. Salzbrenner and Barrick wanted to go even further, creating a way to test N95 masks for reuse—a common practice during the height of the pandemic but for which there is no testing standard.

IMAGE: Sandia National Laboratories

“We developed the chamber version to automate donning and doffing (putting on and taking off an item) to test respirator function over time, a predominant factor in wear on a mask,” Salzbrenner said. “It also mimics how a mask is set on the face and shows you any gaps that air and particles can get past.”

The hope is that these two methods can be combined into an all-in-one tester that accounts for all the aspects of the mask. As Omana explained, the current approach to testing doesn’t account for all the other potential points of failure besides the filter media. This new holistic test gets much closer to how N95 masks are used in the field.

The Sandia team is currently working on further refining their approach, with the goal of licensing the science to a company that can produce their all-in-one testers on a commercial level.

The post A Better Way to Test N95 Respirators appeared first on Engineering.com.

But we shouldn’t let these fancy new versions of predictive text obscure the fact that AI has been around and in development for decades, with numerous productive advances that predate ChatGPT and its ilk. There’s no better way to separate hype from reality when it comes to AI than by taking a look at some of the practical ways it’s being used in manufacturing, and more specifically in quality control (QC).

From machine vision to predictive analytics, here are three ways to use AI for QC. These are presented in descending order of maturity, i.e., the first example is already being deployed in practical applications, the second is mostly still in the pilot phase and the third is, at this point, mostly conceptual.

Unlike LLMs, which have an uncertain future at best, artificial intelligence in quality control has already proven to be productive, with many more applications to grow into. Whether and what effect it will have on the future of the manufacturing workforce remains to be seen but, whatever happens, it’s certainly an interesting time to be a quality professional.

#1 Machine Vision

Machine vision is a form of industrial automation utilized for inspection, sorting and robot guidance. The idea is to use a combination of lighting, cameras and software to extract information from a captured image. This information can be as simple as a go/no-go signal or as complex as the identity, orientation and position of each object in the image.

While machine vision doesn’t involve AI per se, the two technologies are becoming more closely intertwined as developers turn to neural networks in order to augment machine-vision algorithms and improve their accuracy. Audi, for example, has begun using AI for quality control of spot welds at its Neckarsulm plant in Germany. Prior to deploying machine vision, employees had to check the quality of welds manually using ultrasound, with samples drawn randomly.

Audi’s head of production planning, automation and digitization has claimed elsewhere that using machine vision with AI reduced the labor costs associated with these inspections by 30-50 percent. As such, it should come as no surprise that Volkswagen Group is rolling out the technology in Audi’s Brussels factory, the Volkswagen plant in Emden and Audi headquarters in Ingolstadt. All that’s required is to retrain the AI model for the different weld settings in each of these locations.

#2 Root Cause Analysis

The goal of improving manufacturing quality can be fulfilled in two complementary ways: by detecting defects in products and by identifying the source of those defects in the production process. As the example above makes clear, AI can assist in the first task by augmenting machine vision. The second task, more formally known as root cause analysis (RCA), can also benefit from the use of artificial intelligence.

Traditional RCA methodologies include Pareto analysis, fishbone diagrams and the five whys, among others. While these remain essential tools for quality professionals, they also require considerable knowledge and expertise in order to be used most effectively. This presents a problem for manufacturers staring down the barrel of the skills gap, worrying that their best quality people (in both senses of the term) will soon be retiring. When this risk is combined with the fact that manufacturers are producing more data than ever, an opportunity to deploy artificial intelligence as a solution seems only natural.

Here’s the basic idea: take all of your manufacturing process and product data and feed it to a machine learning model (typically, more than one). Through training, the model(s) will eventually recognize the correlates of product defects in the process data (e.g., a variation in the speed or feed of a CNC lathe that correlates with a gear not fitting properly into an assembly). Ideally, this entire procedure will happen in real time, giving manufacturing engineers enough of a heads-up that they can intervene in the process and avoid have to scrap or rework expensive parts.

#3 Supply Chain Management

Precipitated by the COVID-19 pandemic, the ongoing struggles of the global supply chain are continuing to impact manufacturers at virtually every stage of the product lifecycle. Amidst natural disasters brought on by climate change and geopolitical upheavals like Russia’s invasion of Ukraine, the availability of essential parts and materials seems less predictable than ever.

The recent shortage of microprocessors and its effect on the automotive industry is a prime example of how disruptions at one point in the supply chain can bring about serious consequences downstream. However, as with the other examples discussed in this article, managing supply chains is a task to which AI is particularly well suited. The combination of large volumes of data and a need to optimize across multiple parameters is what makes supply chain management such a tractable problem for AI.

By integrating the data from manufacturing execution systems (MES) and enterprise resource planning (ERP), machine learning models can forecast product demand and the availability of raw materials. There are myriad ways this integration can benefit manufacturers, from better data quality to enhancing cybersecurity. However, in the context of manufacturing quality, by putting this information together with supplier performance data, including delivery times and pricing, manufacturers can use AI to ensure consistent product quality across their entire supply chain.

The post How to Use AI for Quality Control in Manufacturing appeared first on Engineering.com.

That level of candor is refreshing in a world where every new piece of software is marketed as a “solution” (often one looking for a problem, rather than the other way round). You won’t hear many global organizations admitting that they have a mess that needs to be cleaned up but, in this case, Zinke’s description is apt.

Hexagon is a sprawling organization, with divisions encompassing everything from agriculture to mining to infrastructure and, of course, product design and manufacturing. That means a lot of data: equipment statuses, building codes, product specifications—all of it scattered across various formats and siloes. But if engineers have learned anything about data, it’s that extracting its full value means being able to leverage it in the right ways by making it available at the right times.

That’s the idea behind Hexagon’s digital reality platform in a nutshell: make data more accessible within and across organizations. And what better place to start than inside Hexagon itself?

According to Zinke, Nexus began as an internal tool for three reasons. “It’s about the gravity of the platform—getting buy-in from other parts of the organization. It also demonstrates commitment, because if you’re using it internally then customers know it’s not just an experiment. And finally, it’s also about maturity: eating your own dog food. If you never use the platform because it’s not mature enough, why would you turn it over to external customers?”

Openness and Transparency

Zinke frames Nexus as being data-centric, designed to be open, transparent and accessible. “If I want to share a PDF with you,” he explains, “I don’t care what you’ll be using to read it and you don’t care what I used to produce it. We think it should be exactly the same with engineering data.”

In practical terms, that means users can read and write data using Nexus without needing a proprietary application programming interface (API). This is epitomized in the recently announced Nexus for Developers, which is intended to accelerate third-party app development for Nexus by providing the software development kit (SDK) building blocks to connect other products and services to Nexus.

The goal is a kind of networking effect: the more products and services that are connected to Nexus (and to each other), the more valuable the platform becomes, not just for Hexagon but for manufacturers as well. Most manufacturing engineers have heard the familiar complaint about data siloes, where the design and production teams practically speak different languages because of the diversity of their toolsets.

Nexus could be the Rosetta Stone for engineering, if it can deliver what Hexagon is promising.

The Limits of Interoperability

For a platform built to connect innumerable engineering workflows, files and formats, ‘interoperability’ is clearly the rallying cry for Nexus. Asked about the limitations of Nexus’s interoperability, Zinke points out that it’s important to distinguish between syntactic interoperability and semantic interoperability.

The former is purely concerned with reading data but this is still a powerful capability on its own, because it enables engineers to do computations, perform statistical analyses and so on.

In contrast, semantic control requires understanding and preserving information across different applications as well as different programming languages. “Semantic interoperability requires working with different bodies that focus on the standardization of data or leveraging existing standards,” Zinke says. “The other aspect of it is ease of integration, which is more on the API/SDK side. We embrace the web ecosystem so that it’s easy to integrate with web apps and cloud apps. These days, web technology can be easily integrated with desktop systems, which is what we do ourselves.”

Hexagon’s approach here is based on open-source technology, specifically Microsoft’s Fluid Framework. This is another example of a development decision that would have been unusual from a large corporation such as Hexagon—or, for that matter, Microsoft—even a few years ago.

However, as Zinke points out, there are clear benefits to taking an open-source approach and avoiding proprietary technology wherever possible: “If you build proprietary tech, you now need to maintain it forever and it’s harder to adopt because your customers need to think about licensing and proprietary contracts. It doesn’t help you scale and it’s also not differentiating.”

What’s Next for Nexus?

While the scope of Nexus is undeniably ambitious, its rollout is proceeding at a more measured pace. This is reflected not only in its early development as an internal product, but in how it’s being delivered to Hexagon’s customers, subsidiaries and partners. “We’re currently in the preview phase, if you will,” says Zinke. “So we’re intentionally restricting the number of users until we get to the pilot phase in Q4 of this year.”

By that time, the Nexus ecosystem will include not only Altium, the electronic design automation software company which was the first to connect its cloud platform, but Nvidia’s Omniverse as well. And with every additional cloud and platform that connects, Nexus becomes that much more useful to designers and engineers, even though they may not realize it on a day-to-day basis.

“The idea is not to disrupt any workflows,” Zinke says, “because many of our customers are using desktop apps, so we don’t want to break anything but just, step-by-step, make those more powerful. In those cases, they’ll just see the new capabilities enabled by Nexus. For developers, there will be those building blocks to connect the data fabric, with visual collaboration, compute capabilities and so on.”

If data is the lifeblood of manufacturing, then Nexus is aiming to become a new circulatory system to invigorate its vital organs—CAD/CAM, PLM, and so on—rather than replace them.

The question is: Will the host accept this transplant?

The post Hexagon’s Nexus Aims to Be The Rosetta Stone for Engineering appeared first on Engineering.com.