China Manufacturing Decoded

Product Integration: Why Working Subsystems Fail When Assembled

Your electronics work. The firmware runs correctly. The mechanical components fit, and every subsystem has passed its individual tests.

Then everything is installed inside the final enclosure, and entirely new problems begin to appear.

In episode 335 of China Manufacturing Decoded, Adrian is joined again by Paul Adams from the group's contract manufacturer, Agilian Technology, to explore one of the most difficult stages of hardware development: product integration.

 

Show Sections

• 00:21 – When Individually Working Parts Fail Together
• 02:28 – Why Everything Works During Bench Testing
• 07:40 – What Changes When Integration Begins
• 08:39 – Designing the Product So It Can Be Debugged
• 10:16 – Integrating One Subsystem at a Time
• 13:45 – A Technically Working Product Can Still Fail
• 15:00 – The Hidden Impact of Heat Buildup
• 20:19 – The First Complete Build Is Only a Mini Milestone
• 21:39 – Solving an Enclosure Airflow Problem
• 23:53 – Paul’s Product-Integration Playbook
• 24:07 – Step 1: Integrate in Stages
• 24:52 – Step 2: Test Important Risks Early
• 26:16 – Step 3: Keep Debugging Access Available
• 27:18 – Step 4: Test the Way the Customer Will Use It
• 28:20 – Step 5: Treat the Enclosure as Part of the System
• 29:29 – Final Lessons From the Integration Process

 

Related content

• How Many Prototypes Are Needed Before We Get ‘Perfection?’
• Transitioning to Manufacturing from Product Development | 2 Options
• Why does new product development take so long?
• Typical Steps Before You Get Look and Work-Like Prototypes
• An Effective New Product Development Process for Electronics
• Final Prototype of your New Hardware Product: When Will You Get There?
• Manufacturing Pilot Runs Are Great With Quantified Objectives
• Handover to Manufacturing: What NOT to do & Best Practices

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

Are Trade Shows Still Worth It? How to Find and Assess Suppliers (Listener Question)

Are trade shows still worth attending when buyers can search for suppliers through Alibaba, Google, ChatGPT, and other online platforms?

For physical products, Renaud Anjoran believes they still offer considerable value, provided buyers arrive with a clear objective and understand how to assess the companies they meet.

In this listener Q&A episode, Adrian and Renaud answer Robert C’s question about events such as the Canton Fair, Global Sources, CES, and IFA.

 

Show Sections

00:00:10 – Introduction and Robert’s listener question

00:01:19 – Are trade shows still worth attending?

00:03:40 – Why in-person trade shows can beat online supplier searches

00:04:53 – Spotting trends and avoiding undifferentiated products

00:07:18 – Why an impressive booth proves very little

00:09:06 – Questions buyers should ask potential suppliers

00:09:32 – How to avoid being seen as a “tire kicker”

00:11:35 – Researching exhibitors before attending the show

00:14:11 – Manufacturer or trading company?

00:15:00 – In-house capabilities, customer fit, and production capacity

00:16:52 – How suppliers judge visitors at their booths

00:18:21 – How to present your company as a credible prospect

00:20:12 – Realistic forecasts, order values, and exaggerated promises

00:21:48 – Why the maturity of your product design matters

00:23:16 – Discussing first-order MOQs and longer-term volumes

00:24:39 – Final advice and the Agilian factory-tour series - watch the tour videos here

 

Related content

• How To Get More Out of a China Trade Fair Visit
• How To Find Suppliers in China
• 27 Questions To Ask During a China Factory Visit
• Are Suppliers on Alibaba and Global Sources Trustworthy?
• Sourcing from China 101, Part 2: How to Identify Potential Chinese Suppliers?

Found some possible suppliers? We can help check them:

• Supplier Legal Records Check
• Certificates & Reports Verification
• Supplier Bank Account Verification

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

Gold: 3 QC Plans You Need To Make Before Production (Ep. 30 Revisited)

Adrian revisits a classic episode (originally Ep. 30, Nov 2020) in this monthly rewind, Episode 333 of China Manufacturing Decoded, and sits down with Sofeast CEO, Renaud, to unpack the three types of quality-control (QC) plans every importer and manufacturer should agree before production starts.

Renaud explains three types of QC plans:

1. The product quality control plan, often linked to the manufacturing contract.
2. The process control plan, which controls quality during production rather than waiting until the end.
3. The QC plan for new products, which helps buyers and suppliers think through risks before mass production starts.

Key takeaways for listeners: QC planning is a pre-production activity not a rescue job; define a clear product quality standard; decide how quality will be controlled during production; and for any new product, agree on what must be proven before mass production. If any of those points are unclear, that’s likely where your next quality risk is hiding.

 

Show Sections

• 00:00 Introduction to this rewind episode
• 01:56 What quality control plans are and why they are needed before production
• 04:06 Why there is more than one type of QC plan
• 04:43 Type 1: The product QC plan and contract-related quality terms
• 05:28 Defining testing, inspections, AQL limits, compliance, and responsibilities
• 06:41 What happens if serious issues are found after shipment?
• 07:11 Why even smaller buyers should document quality expectations
• 08:06 Type 2: The process control plan
• 09:04 Mapping production processes and critical steps
• 10:20 Turning the control plan into work instructions and checks
• 11:02 When process control plans become important
• 11:54 Why final inspection alone is often too late
• 12:27 Controlling quality through incoming components and sub-suppliers
• 13:50 How to check whether suppliers can follow process control plans
• 15:03 Type 3: The QC plan for a new product
• 16:27 Quality, reliability, and compliance requirements
• 17:35 Golden samples and approved prototypes
• 18:00 Testing stations, jigs, fixtures, and functional checks
• 19:07 Intended use, reliability expectations, and compliance needs
• 19:52 Component manufacturing, assembly, tooling, and work instructions
• 21:35 Pilot runs and pre-production approvals
• 22:35 Why new products force buyers and suppliers to think harder
• 22:59 Supplier optimism and the “we’ll fix it later” risk
• 24:16 Why quality standards need to be clear and useful
• 25:08 Why buyers often skip proper QC planning
• 26:42 Why defining requirements is the buyer’s job
• 27:40 Which QC plans apply to which buyers and products?
• 28:22 QC planning for all buyers vs larger or higher-risk buyers
• 29:26 Why process control is worth considering for new products
• 30:12 Why every buyer still needs at least a basic quality standard
• 31:12 What off-the-shelf and private-label buyers should focus on
• 33:02 2026 outro and key lesson recap

 

Related content

• Quality Control Plan: Defining Expectations Before Production
• How To Set Up A Process Control Plan [11 Steps]
• Golden Sample in Manufacturing
• What Is A PP Sample?
• How to set product specifications?
• You NEED to do product qualification BEFORE mass production!
• How Incoming Quality Control Inspections Fit into an Overall Quality System

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

Setting Up a New Factory? Ask These Questions First (Feat. David Collins III, CEO of MTG)

Setting up a new factory is a major strategic decision. It is not just about finding cheaper land, moving away from China, or following other companies into Vietnam, Mexico, or another popular manufacturing location.

In this episode, Renaud speaks with David Collins, CEO of Manufacturing Transformation Group, about what companies need to think through before relocating production or building their own factory.

They discuss why more companies are considering factory relocation or ownership again, especially after COVID, tariff changes, supplier dependency, and IP concerns. But David explains why the first question should not be “where should we move?” It should be “what are we actually trying to accomplish?”

The conversation covers the real trade-offs between China, Vietnam, Mexico, and other locations; why labour cost should not be the only driver; how supplier location, workforce skills, logistics, and infrastructure affect the decision; and why companies need a proper BOM, cost model, and feasibility study before making a move.

They also get into greenfield vs brownfield factory projects, equipment selection, factory layout, commissioning, factory acceptance testing, and why automation can be a waste of money if it does not fit the real production process.

The key message: moving to a new factory is a rare chance to redesign your manufacturing system properly. But if you simply copy the same poor layout, weak supply chain, bad inventory habits, and unsuitable equipment into a new building, you may just move the mess.

 

Show Sections

00:00 – Introduction: setting up a new factory
01:43 – Who David Collins and Manufacturing Transformation Group are
05:04 – Why more companies are considering factory relocation
05:50 – China, Vietnam, Mexico, and the real trade-offs between locations
08:10 – Why some companies want to own manufacturing again
09:32 – Don’t just move the mess to a new factory
11:45 – The first question: what are you trying to accomplish?
12:02 – Supplier location, workforce skills, logistics, and infrastructure
14:18 – Why a real BOM and cost model are essential
15:27 – Feasibility studies and idealised factory planning
16:07 – Why automation is not always the right answer
17:34 – Comparing factory setup scenarios and locations
18:16 – Why labour cost should not be the only driver
20:48 – IP risks and supplier dependency
22:15 – Learning from the problems in your current factory
23:46 – Project management during a factory move
24:03 – Greenfield vs brownfield factory projects
26:09 – Layout planning, implementation, and local specialists
27:13 – On-the-ground project management and construction risks
28:33 – Equipment commissioning and factory acceptance testing
29:50 – Choosing equipment that fits your real needs
31:41 – Equipment maintenance, spare parts, and supplier risks
32:40 – Why factory setup is a once-in-a-decade decision
34:12 – Disciplined planning and avoiding old mistakes
36:45 – Closing thoughts

 

Related content

• How To Plan for Transferring Production To a New Factory: 45 Point Checklist
• Transfer Manufacturing From One Chinese Factory To Another With Fewer Risks
• How To Diversify Manufacturing Sources Out of China and Cut Risk
• Sofeast can help you > Electronic Production Transfer from China to India OR Malaysia
• Supply Chain Risk Management, Part 5: Moving Manufacturing to Vietnam, Thailand, Malaysia, or India (Pros & Cons)
• Production Transfer: A Roadmap (Assembly Operations Only)

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

The Truth Behind “8–12 Weeks”: Injection Mold Tooling Timelines Exposed

In this episode of China Manufacturing Decoded, Adrian is joined by Paul Adams, Head of New Product Development at the Sofeast Group's contract manufacturer Agilian Technology, to discuss one of the most common assumptions hardware founders make before moving into tooling: that tooling will take “8 to 12 weeks.”

Paul explains why that figure can be true in very simple cases, but why it is often misleading for real consumer electronics, IoT, and hardware products. Tooling timelines depend on design readiness, DFM review, part complexity, steel selection, toolmaker capacity, customer responsiveness, and the timing of Chinese holidays such as Chinese New Year and Golden Week.

They also discuss why the tooling clock does not really start when the purchase order is placed, why T0, T1, and T2 trials need to be planned carefully, and why founders should build schedule buffers before cutting steel.

For hardware startups and product teams preparing for injection molding, metal stamping, die casting, or other production tooling, this episode explains how to build a more realistic tooling schedule and avoid costly launch delays.

 

Podcast sections

• 00:00:31 – The “8 to 12 week tooling timeline”
• 00:02:28 – What tooling includes and why it matters
• 00:04:21 – Tooling cost and why first-time founders get caught out
• 00:06:08 – Where the 8 to 12 week figure comes from
• 00:07:23 – Why real consumer electronics products are more complex
• 00:08:35 – When the tooling timer really starts
• 00:11:10 – Why design readiness and DFM review are critical
• 00:13:26 – How part complexity affects tooling lead time
• 00:13:50 – Steel selection: P20, H13, and tool life
• 00:15:40 – Responsiveness during T0, T1, and T2 trials
• 00:16:26 – Why being in China can speed up tooling decisions
• 00:19:03 – Planning around Chinese New Year, Golden Week, and May Day
• 00:21:47 – How to create a tooling schedule that works
• 00:22:05 – Reviewing the DFM report properly before cutting steel
• 00:24:00 – Building a tooling specification and critical path plan
• 00:25:34 – Understanding T0, T1, T2, and rework cycles
• 00:27:45 – Why you should always build in a schedule buffer
• 00:28:56 – Why many tooling delays come from the customer side
• 00:30:15 – Final advice: understand the full tooling process

 

Related content

• Tooling Management for Plastic Injection Molds in China
• Plastic Injection Mold Tooling Management & Risk Reduction [Podcast]
• Common Design For Manufacture Improvements On Plastic Injection Molded Parts
• Injection Mold Tooling Roadmap: How to Get from Smart Design to T1 Samples
• What are Plastic Injection Mold Tooling Revisions? (3 examples)
• How To Make Faster Injection Mold Tooling [7 Tips]
• Plastic Injection Molding Pilot Runs: What You Need To Know
• The Four Levels of Plastic Injection Molding Suppliers in China

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

QC During NPI: Build Quality In Before Mass Production

Episode 330 of China Manufacturing Decoded features hosts Adrian and Renaud from the Sofeast Group discussing why quality control should not start when finished products come off the production line. By then, many key decisions have already been made: product requirements, supplier selection, component choices, tooling, process setup, inspection methods, and testing plans.

In this episode, Adrian and Renaud explain what quality control should look like during the NPI process, before mass production begins. They discuss why final inspection is only one part of the quality picture, how clear product requirements reduce confusion, why supplier and component qualification matter, and how process controls, inspection points, test methods, jigs, fixtures, and pilot runs help prevent defects before they become expensive production problems.

You’ll learn why quality needs to be built into the product and manufacturing process from the start, rather than being inspected in at the end.

The main takeaway: final inspection may catch problems, but it does not prevent them. Good NPI quality control reduces risk earlier, when changes are easier and cheaper to make.

 

Podcast sections

• 00:00:11 Episode 330 begins: QC during NPI before mass production
• 00:01:14 Why many companies treat quality control as an end-of-line activity
• 00:02:08 Why final inspection is reactive, not preventive
• 00:04:01 How to build quality into the product and process earlier
• 00:04:44 Why everything in product development can affect quality
• 00:06:08 Product requirements as the foundation of NPI quality control
• 00:07:09 Supplier qualification, design risks, inspection, and testing
• 00:08:29 Quality gates, validation, reliability, compliance, and performance
• 00:09:36 Manufacturing process controls and why they need to be planned
• 00:12:02 Using AI to help document product requirements
• 00:13:00 Examples of turning user needs into measurable specifications
• 00:15:41 Cosmetic standards, boundary samples, and critical measurements
• 00:18:21 Qualifying suppliers, components, and materials
• 00:19:53 Turning requirements into inspection and testing processes
• 00:22:18 Applying QC controls during prototype and pilot batches
• 00:23:04 Work instructions, jigs, fixtures, and process risk reviews
• 00:25:05 Mistake proofing example: preventing drilling errors
• 00:26:28 Eliminating risks where possible, controlling them where not
• 00:27:12 Why prevention is stronger than end-of-line inspection
• 00:28:04 Final takeaway: quality-forward NPI reduces production risk

 

Related content

• NPI process guide
• NPI deliverables review service from Sofeast
• 7 must-do NPI tasks before a successful launch
• Why skipping part qualification in NPI will cause problems
• 3 key process improvement tools
• Pilot run best practices
• DFM and Industrialization support from Agilian
• You NEED to do product qualification BEFORE mass production!

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

Why Working Prototypes Fail in Production, Part 2: The Failure Patterns and Fixes

Why do some working prototypes still fail when they reach production?

This is episode 329 and the second part of our discussion on this topic, and Adrian and Paul move from the general prototype-to-production gap into real-world failure patterns that can derail a product launch. They look at 3 common scenarios:

1. Component swaps made for cost reduction
2. Firmware clean-up before release
3. And transferring production from one factory to another

You’ll hear why a cheaper component that looks identical on paper can still cause major problems, why every firmware change needs to be tested and documented, and why a factory transfer should never be treated as a simple handover.

The episode also explains how a structured NPI/MPI process, production-representative builds, configuration control, phase gates, pilot runs, and factory process audits help reduce the risk of production failure.

The key message: a prototype proves the concept, but production proves the process. Before approving production, you need to know exactly what was validated, what configuration it applied to, and what has changed since.

 

TIMESTAMPS

• 00:00 - Introduction: why working prototypes still fail in production
• 01:32 - Failure pattern 1: component swaps and hidden validation risks
• 06:26 - Failure pattern 2: firmware tidy-up before production release
• 08:53 - Failure pattern 3: transferring from prototype shop to production factory
• 13:20 - How to bridge the prototype-to-production gap
• 13:48 - Why a structured NPI process matters
• 14:51 - Production-representative builds, EVT, DVT, tooling, and PVT
• 16:49 - Controlled ramp-up instead of jumping straight to mass production
• 17:32 - Configuration control: validation only applies to what was tested
• 20:29 - Practical decision framework for managers
• 22:03 - Setting a configuration baseline from DVT onward
• 23:05 - Using NPI phase gates and change assessment before moving forward
• 24:29 - Factory process audits: why an audit is not just a factory tour
• 27:09 - Pro tips: quality standards, NPI discipline, and validation tracking
• 30:39 - Factory transfers and why pilot runs are essential
• 33:05 - Final recap: what changed, what was validated, and what is now unknown

 

Related content

Get help with your project from Sofeast. These services cover the topics discussed today:

• New Product Introduction Support
• NPI Deliverables Review
• DFM Review for Manufacturing in Asia
• Reliability Engineering & Testing
• Process Management Audit (PMA)
• First Article Inspection

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

Why Working Prototypes Fail in Production, Part 1: What Changes Before Mass Production

A prototype works. The team signs it off. Everyone feels confident.

Then production starts, and unexpected failures appear.

Why does this happen?

In this episode, Adrian is joined by Paul Adams, the Sofeast Group's Head of New Product Development, to discuss the gap between prototype and production. This is part one of a two-part discussion on why working prototypes can still fail once products move toward mass production.

Paul explains why prototypes and production units are often not the same thing, even when they look identical. The episode covers five areas where important changes can creep in:

1. Components
2. Firmware
3. Suppliers and factories
4. Tolerances and process variation
5. Validation basis

The key point is simple:

A prototype proves the concept. Production proves the process.

Understanding that difference helps hardware teams, product developers, and importers avoid painful surprises when moving from a successful prototype to production.

In part two, next week, we’ll continue the discussion by looking at common real-world failure patterns, including component swaps, firmware tidy-ups, factory transfers, and how a structured NPI process helps close the gap.

 

TIMESTAMPS

• 00:00 Introduction: why working prototypes can still fail
• 02:09 Prototypes and production units are not the same thing
• 03:46 The gap between prototype and production
• 04:23 Five things that change before production
• 04:36 1 - Components: prototype parts vs production parts
• 09:17 2 - Firmware: why prototype code is not production-ready
• 12:03 3 - Suppliers and factories: why process knowledge gets lost
• 16:50 4 - Tolerances and process variation
• 19:54 5 - Validation basis: What exactly was tested?
• 22:22 Key takeaway from part one
• 23:17 What to expect in part two

 

Related content

• How Many Prototypes Are Needed Before We Get ‘Perfection?’
• Process Management Audit (PMA)
• An Effective New Product Development Process for Electronics
• From Prototype to Production: 7 Pitfalls for Tech Products

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

Gold: NRE Costs Exposed: How One-Time Engineering Bills Can Sink Your Product (Ep. 49 revisited)

Host Adrian revisits episode 49 (a ‘gold episode’ originally recorded in 2021), a topic that still catches many product developers and importers by surprise: non-recurring engineering costs, often shortened to NRE costs.

These are the one-time costs needed to get a new product ready for production, such as engineering work, product design, prototyping, tooling, supplier sourcing, reliability testing, compliance testing, testing fixtures, and production setup.

If you underestimate NRE costs, your product plan may look profitable on paper but fall apart before launch. This episode explains what NRE costs are, why they can grow quickly, where they appear in different manufacturing processes, and how to protect yourself with better planning, supplier due diligence, and the right development agreements.

 

TIMESTAMPS

• 00:00 — Intro: why NRE costs still matter
• 01:13 — What are non-recurring engineering costs?
• 03:04 — Why NRE costs affect your real product margin
• 04:16 — Why NRE budgets often grow during development
• 07:37 — Typical NRE costs by product and manufacturing process
• 08:10 — Plastic injection molding and tooling costs
• 10:44 — Custom PCBAs and electronics engineering costs
• 13:46 — Why NRE planning affects cost and delivery time
• 15:53 — Existing tooling, white-label products, and off-the-shelf options
• 18:51 — IP and dependency risks with ODM products
• 20:08 — When a manufacturer offers to absorb NRE costs
• 22:03 — Why a development agreement matters
• 24:27 — Why manufacturers prefer production over development work
• 26:39 — A working prototype does not mean you are production-ready
• 29:04 — Final summary: what to include in your NRE planning

 

Related content

• What is an NRE Cost (Non-Recurring Engineering)?
• Costs and Milestones to go from Product Concept to Market?
• How to Cost Your Product Properly (Design-to-Cost Explained)

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB

Why Hardware Projects Stall: Avoiding 'Failure to Launch'

In episode 236, we explore why so many hardware products never make it to market, even when the idea is strong, the team is ready, and the budget is there.

In this episode of China Manufacturing Decoded, your host Adrian is joined by Paul Adams from Agilian, part of the Sofeast Group, to break down the real reasons hardware projects stall before they even start, and what you can do to avoid it.

They go beyond theory and share practical lessons from real projects, including costly mistakes around missing specifications, bad assumptions, and external pressure to move too fast.

You’ll learn:

• Why missing product requirements quietly kill projects
• The difference between having an idea and being ready to start
• How assumptions compound into expensive errors
• The hidden risks in BOMs, components, and compliance
• Real-world case studies where projects stalled, and why
• A practical 10-point checklist to validate your readiness before development

The goal of this episode is to help you avoid delays, wasted budget, and failed launches when you're launching your product.

🎧 Listen now and make sure your next product is built on solid ground.

 

TIMESTAMPS

• 00:03 — Intro & episode overview
• 01:01 — The “failure to launch” problem in hardware
• 02:01 — It’s not the team: real root causes
• 03:02 — Assumptions & missing information (core issue)
• 07:00 — Red flags: missing requirements & BOM
• 11:57 — What “ready to start” actually means
• 12:45 — NPI process & phase gates explained
• 14:22 — Specs as a living document (market changes
• 15:05 — Mechanical, electronics & feature requirements
• 17:34 — Volume assumptions & pricing impact
• 19:08 — The danger of rushing decisions
• 20:44 — Case study: prototyping failure under pressure
• 24:25 — Case study: component & supply chain risks
• 26:33 — Case study: regulatory & certification surprises
• 29:45 — The 10-point pre-start checklist
• 32:53 — Most common mistake
• 33:47 — Final takeaway

 

Related content

• Transitioning to Manufacturing from Product Development | 2 Options
• IP Protection in China when Developing Your New Product [Importer’s Guide]
• Bill of Materials (BoM) Explained
• Design to Cost (DTC) Explained
• Getting To Grips With Non-Recurring Engineering Costs (NRE) [Podcast]
• 11 Common Electronic Product Certification And Compliance Requirements
• Crowdfunding Failures: 4 Great Prototypes That Failed To Launch
• Learn more about how we handle DFM & Industrialization (NPI) for our manufacturing customers

Get in touch with us

• • Connect with us on LinkedIn
  • Contact us via Sofeast's contact page
  • Subscribe to our YouTube channel
  • Prefer Facebook? Check us out on FB