China Manufacturing Decoded

Episode 350 of China Manufacturing Decoded features host Adrian and NPI lead Paul Adams as they unpack a common but overlooked quality risk: you can’t always trust your factory's measurements. They explain why identical parts can show different readings depending on equipment, operators, methods and environmental conditions, and why a single “it’s within tolerance” result may be misleading.

Paul explains how the resolution and calibration limits of measuring tools matter (from steel rules and digital callipers to coordinate measuring machines), why measurement methodology and handling must be standardised, and how temperature, cleanliness, and operator technique influence results. He emphasizes the need for written SOPs, qualified inspectors and daily or periodic calibration checks with traceable records.

The episode introduces Gage (Gauge) R&R (repeatability & reproducibility), the structured test that has multiple operators measure the same parts with the same device to quantify variability, and describes a real-case study involving an unreliable air-pressure test rig. That case study shows how flawed measurement systems can create false failures or allow escapes, and how fixing the test equipment and repeating the Gage R&R restored confidence in results.

They also cover statistical process control and Cpk: how control charts and capability analysis reveal when a machining process is stable, trending out of spec, or consistently off-target (the bullseye vs. tight-but-offset metaphor). Paul stresses focusing inspection effort on critical-to-quality (CTQ) dimensions rather than every nonessential measurement.

Key takeaways for buyers and engineers: don’t blindly accept measurement reports; ask for calibration records, written measurement methods, evidence of Gage R&R and SPC on CTQs, and confirmation that inspectors are trained. The episode ends with practical advice on preventing defects by building reliable measurement systems early in development and production, plus where to seek help from Sofeast for inspections, auditing and manufacturing support in Asia.

Topics discussed include:

Why a measurement reading should not automatically be treated as fact

How the measuring equipment, operator, environment, and technique can all affect results

Why tighter tolerances require more capable measurement equipment

Why calibration records matter when assessing a supplier’s quality system

Why factories should use clearly defined measurement procedures

What Gage R&R is and how it helps determine whether measurements are repeatable

How unreliable test equipment can make good products appear defective

Why poor measurement systems can lead to both false rejects and quality escapes

How Cp, Cpk, control charts, and statistical process control help monitor manufacturing processes

Why manufacturers should focus these controls on Critical to Quality (CTQ) characteristics rather than attempting to control every dimension equally

What buyers should ask suppliers before accepting their measurement reports at face value

This episode is brought to you by the Sofeast Group. Visit sofeast.com for inspection, NPI support, contract manufacturing, 3PL, and other services across Asia’s manufacturing hubs.

 

Get in touch with Sofeast Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

Dyson spent six years developing its $500 CameraJet toothbrush, involving hundreds of engineers and some highly sophisticated technology.

But within days of its September 2026 launch, users and reviewers were reporting failures. The product was soon temporarily unavailable, and Dyson later confirmed that one production batch of jet tubes had experienced a manufacturing process issue.

So how can a product with so much engineering behind it still run into problems almost immediately after launch?

In this episode, Adrian and Renaud use the Dyson CameraJet story as a real-world case study in the risks hardware companies face when moving from product development into mass production.

They discuss why waterproof products are particularly difficult to manufacture reliably, how a good design can still be undermined by process variation, and why successful prototypes and validation testing do not necessarily prove that a production line is ready.

The conversation also looks at the importance of a structured NPI process, production validation builds, end-of-line testing, process control, and involving experienced manufacturers early enough to uncover problems before a major commercial launch.

Finally, they discuss what should happen when failures start appearing in the field: gathering failed units quickly, identifying patterns, performing failure analysis, feeding lessons back into design and manufacturing, and using traceability to isolate affected batches.

Topics covered include:

What happened during the Dyson CameraJet launch

Why waterproofing is such a difficult hardware challenge

How manufacturing variation can undermine a sound product design

Why early mass production is often the highest-risk period

The role of EVT, DVT and production validation

Why test fixtures and end-of-line testing need to mature before launch

Process control during production ramp-up

The risks of a high-profile “big bang” product launch

How to investigate early field failures

Why manufacturing traceability matters when one batch has a problem

What SMEs and hardware startups can learn from Dyson’s experience

This episode is brought to you by the Sofeast Group. Visit sofeast.com for inspection, NPI support, contract manufacturing, 3PL, and other services across Asia’s manufacturing hubs.

Get in touch with Sofeast Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

Rushing a new hardware product into mass production can feel like the fastest route to market. But both buyers and suppliers have commercial reasons to skip NPI steps, and those shortcuts can create much more expensive problems later.

In episode 348 of China Manufacturing Decoded, Adrian and Renaud look at why manufacturers often prefer to move into mass production, where they make their money, rather than spend months on engineering work, small pilot runs, process development, and repeated validation.

They explain when moving quickly can be relatively low-risk for a familiar, established product, and why the same approach becomes dangerous when a product is new, highly customized, or requires new manufacturing and testing processes.

The buyer side matters too. Launch deadlines, promises to customers, funding constraints, cash-flow pressure, and the need to start generating revenue can all tempt companies to commit to large production quantities before the design and manufacturing process are truly ready.

The result can be custom components, tooling, labor, and thousands of finished products already committed before a serious problem is discovered.

Ultimately, the better question isn't simply, “How quickly can we start mass production?” It's: “How quickly can we reach mass production that is predictable and consistent?”

 

Topics Discussed

Why Chinese manufacturers are commercially motivated to reach mass production

When moving quickly into production may carry relatively low risk

Why new and customized products need their own manufacturing and testing processes

How to assess whether a supplier is really capable of supporting NPI

The dangerous assumption that a manufacturer of similar products can automatically develop yours

Why buyers also knowingly skip NPI steps

How deadlines, customer commitments, and funding pressure influence decisions

The financial risk of committing components and materials too early

What happens when problems appear after thousands of units have already been produced

Why progressive pilot runs reduce manufacturing risk

The difference between reaching mass production quickly and reaching stable mass production

When adapting an existing product can reduce development and industrialization risk

 

Related content

Sofeast – Transitioning to Manufacturing from Product Development | 2 Options

Sofeast – The NPI Process: Trouble Awaits If You Skip Its Steps!

Sofeast – Planning For Effective Pilot Runs

Sofeast – Discussing the NPI Process for Effective Product Development

Get in touch with Sofeast Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

Poor communication can quietly derail a hardware project long before anyone realizes there is a problem.

In episode 347 of China Manufacturing Decoded, Adrian and Paul investigate how information moves from the customer’s Product Requirements Document through design, engineering, purchasing, production, and suppliers, and how misunderstandings or unspoken assumptions at any of those handoffs can eventually turn into expensive hardware problems.

Unlike software, where an error can often be corrected relatively quickly, a misunderstanding in a hardware project might not become obvious until a prototype has been built, tooling has been made, components have been ordered, or production is already underway.

They'll discuss why communication needs to be a two-way process rather than simply passing instructions downstream. Teams need to question unclear requirements, confirm what they have understood, expose assumptions, and send questions back upstream before work continues.

Finally, they also look at the problems caused by changes during development and pre-production. Even seemingly minor changes can affect tooling, components, inventory, cost, lead times, and scrap, making formal change control essential.

 

Topics Discussed

Why hardware projects are particularly vulnerable to communication problems

How assumptions creep in as requirements move between teams

Why simply acknowledging receipt of a requirement is not enough

How a Product Requirements Document should act as the project's single source of truth

Why technical terminology, units, and language differences can cause costly misunderstandings

How two-way requirements reviews help expose problems earlier

Why suppliers should question unclear requirements rather than blindly following them

Which internal teams should be involved in regular project communication

Why mid-project engineering changes can have major downstream consequences

How formal change control protects cost, quality, and delivery schedules

 

Related content

Sofeast - Importing from China? Use this product specifications template.

Sofeast - Good and Bad Methods of Communication with your Chinese Supplier [Podcast]

Sofeast - Why Hardware Projects Stall: Avoiding ‘Failure to Launch’ [Podcast]

QualityInspection.org – Product Requirements Document: Underpinning Hardware Product Development

 

Get in touch with Sofeast Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB

What should you really be designing a new product for?

In this CMD Gold episode, we revisit episode 99, where Renaud Anjoran and reliability expert Andrew Amirnovin explore Design for X (DfX), the practice of designing a product around specific objectives that go far beyond simply making it function.

They examine 12 areas where products can be deliberately optimized, including shorter development times, crowdfunding, manufacturing, assembly, quality, packaging, reliability, maintainability, ergonomics, fewer SKUs, and sustainability.

Along the way, they explain why simplifying products, reducing part counts, using proven components, mistake-proofing assembly, considering the real-world use environment, and involving the right specialists early can prevent costly problems later.

The key message is that DfX is not about maximizing every possible objective. Different products have different priorities, and some objectives involve trade-offs. What matters is identifying the right ones early enough in development that you still have the freedom to make changes.

 

Show Sections

00:13 – Introduction: Revisiting Design for X

01:46 – Design for shorter development and time to market

04:26 – Designing for crowdfunding

07:12 – Design for Manufacturing and Design for Assembly

11:43 – What should product designers consider from the start?

15:57 – Fewer parts, standard components, and mistake-proofing

19:23 – Modularization and simplifying assembly

23:35 – Design for quality

30:52 – Why packaging must be considered during product design

34:06 – Design for reliability

42:58 – Design for maintainability

46:48 – Design for ease of use and ergonomics

47:07 – Design for fewer SKUs

47:53 – Design for sustainability

48:41 – Which DfX objectives matter most?

49:24 – How to implement Design for X in practice

50:53 – Using expert design reviews

51:23 – Adrian's present-day takeaways

 

Related content

Agilian - DfX, Industrialization & NPI Support

Sofeast - The New Product Introduction Process Guide

QualityInspection.org – How Good DFM and DFA Help with Production Cost & Quality

QualityInspection.org – The Design for Reliability Process for Launching Reliable Products

QualityInspection.org – Don’t Leave Packaging Until the End

Agilian – Why Test Components and New Products to Failure During NPI?

Get in touch with Sofeast Connect with us on LinkedIn

Contact us via Sofeast's contact page

Subscribe to our YouTube channel

Prefer Facebook? Check us out on FB