A technically complete product is not necessarily an OEM Manufacturing Project Ready for Production.
For an OEM moving an electronics product from prototype validation to manufacturing, many of the important decisions happen before the first production batch is released.
The BOM may already be complete. The PCB may already have passed prototype testing. The mechanical design may already be finished.
Yet questions can still remain around:
These questions do not necessarily mean that the product design has problems.
They appear naturally when a product moves from engineering validation into manufacturing.
A recent project from a UK customer provides a good example.
The product was a 4G data logger designed for rally cars. It reads vehicle data through the OBD interface and transmits the data through a 4G connection together with location information.
The prototype had already been tested in the UK.
The customer’s next step was to move production to China.
The project eventually moved into production smoothly, but there was no single difficult manufacturing problem that determined the outcome.
Instead, several smaller manufacturing questions were addressed before they became production problems.
The customer initially contacted us with a relatively complete package.
The technical information included the PCBA documentation, while the email itself explained the product application and project background in considerable detail.
The customer also provided:
This allowed us to move the technical documents to purchasing for evaluation quite quickly.
There was little need to spend additional time trying to determine what the product was, where the project stood, or whether the customer was still at the concept stage.
The information was already there.
This matters because an EMS quotation is not simply a calculation of component prices.
Before purchasing starts evaluating the BOM, the manufacturing team needs to understand what is actually being quoted.
This is also one reason why two EMS suppliers can sometimes return noticeably different quotations for the same BOM and Gerber files. The difference may come from the assumptions made during the quotation process.
We discussed this in more detail in our previous article, Why EMS Quotations Are Different.
In this case, the project background was clear enough that the quotation could move forward without a long preliminary investigation.
The initial quotation created an interesting discussion.
The customer told us that our quotation was approximately 8% higher than another supplier’s quotation.
He then asked whether we had used any alternative components.
We had not.
The quotation was based on the customer’s original BOM without unauthorized component substitutions.
The customer was actually pleased with this answer.
For a product that had already been prototyped and tested in the UK, maintaining component consistency was more important than automatically pursuing the lowest possible material cost.
This is an important distinction in OEM manufacturing.
Cost reduction and unauthorized substitution are not the same thing.
A customer may eventually decide to approve a lower-cost alternative, but that decision should remain under the customer’s engineering and purchasing control.
An EMS supplier should not silently change the product in order to make its quotation look more competitive.
We also included pricing references for 1,000 pcs and 5,000 pcs in addition to the initial 100-piece quantity.
The customer appreciated having visibility into how the unit price could change as production volume increased.
This gave him a better basis for evaluating the future manufacturing economics of the product rather than looking only at the initial 100-piece batch.
Before placing the sample order, the customer requested a video meeting.
He did not have enough time to travel to China for an on-site visit, so the video meeting became an important part of the supplier evaluation.
The questions were very practical.
They were not simply:
How many SMT lines do you have?
Instead, the customer wanted to understand whether the supplier could support the product as its manufacturing requirements evolved.
The customer explained that a metal enclosure might be required in the future.
The questions included:
We showed examples of industrial control products with metal enclosures and demonstrated the corresponding waterproofing capability.
The value of this discussion was not simply proving that metal enclosure manufacturing was theoretically possible.
It was about showing experience with similar industrial products and their actual manufacturing requirements.
For an OEM, there is a difference between a capability listed on a supplier’s website and a supplier being able to demonstrate how that capability has been applied to real products.
The customer also asked about multilayer PCB manufacturing and whether we were familiar with via-in-pad processes.
More importantly, he wanted to understand how BGA quality would be controlled during mass production.
It would have been easy to answer:
Yes, we have X-Ray inspection.
But X-Ray is only one part of BGA quality control.
We discussed the soldering process itself, including solder paste selection and handling.
For fine-pitch and BGA assembly, solder paste condition can have a significant impact on solder joint quality.
Factors such as:
all influence the stability of the printing and reflow process.
Poorly handled solder paste can contribute to excessive voiding and other soldering defects.
We showed the customer X-Ray images from previous production, including examples of both unacceptable and acceptable BGA solder joints.
This made the discussion much more practical than simply listing inspection equipment.
The important question in mass production is not only:
Do you have X-Ray?
It is:
How do you control the process that produces the solder joint, and how do you verify the result?
For industrial electronics, process control and inspection work together.
The next discussion involved the 4G communication module.
The customer had selected a Quectel module for the data logger and wanted to understand our relationship with the manufacturer and our ability to source the module.
This is particularly relevant for products where a communication module is a major part of the BOM.
We explained that Fumax already had several mass-production projects using Quectel products and could purchase through established supply channels.
For a larger production program, direct manufacturer sourcing could be considered.
For a smaller initial quantity, such as the 100-piece batch in this project, an authorized distributor could also be used to obtain genuine components in the required quantity.
The customer also had the option of using his own supply channel.
Our quotation included material-level cost details, so the customer could see the purchasing basis rather than receiving only a single unexplained material total.
If the customer had a better purchasing channel, we could use that channel.
Alternatively, the customer could authorize the purchasing relationship through an LOA arrangement if doing so provided better pricing.
This gives an OEM several ways to optimize the supply chain without forcing the entire project into one purchasing model.
For long-term production, this flexibility can become important.
A component may be technically suitable today, but manufacturing readiness also depends on whether it can be sourced consistently, legitimately, and at a reasonable cost as volume changes.
The most interesting manufacturing discussion came after the project entered production preparation.
The customer’s original functional test approach used the actual production firmware and simulated the real operating environment.
From a product validation perspective, this made sense.
However, reproducing the real environment for every production unit created several practical problems.
The test setup required an ECU.
A suitable ECU would cost approximately USD 200 per unit.
Testing every finished unit this way would also be relatively slow and inefficient.
More importantly, we asked the customer another question:
If a defective unit passed through production, could your team repair it after receiving it?
The customer said they did not have their own repair capability.
That changed the risk calculation.
If an intermittent or functional defect escaped the production line, the problem would not simply be a normal manufacturing rework issue.
The unit would have to be returned to the customer, where repair capability was limited.
Therefore, the objective should not simply be to reproduce the customer’s real-world environment during every test.
The objective should be to create a repeatable production test that provides high confidence before the product leaves the factory.
We suggested using two firmware versions.
This firmware would be designed specifically for manufacturing functional testing.
It would allow the test fixture to verify the required functions efficiently without depending on the customer’s expensive ECU environment for every unit.
The production process could then perform 100% functional testing.
After the unit passed the production test, the production firmware would be replaced with the actual firmware intended for the end product.
A sampling check could then be performed using the customer’s actual firmware.
The proposed sequence was:
Production firmware → 100% functional test → pass → customer firmware → sampling verification
We also explained the basic operating principle of the test fixture so that the customer could understand how the production test would work.
The customer agreed with the approach.
This was not a change to the product design.
It was a manufacturing-process decision based on the actual production and after-sales environment.
The original test method was useful for product validation.
The production method simply needed to be optimized for repeatability, efficiency, and risk control.
The project eventually moved through sampling and into the mass-production process smoothly.
Looking back, there was no single manufacturing problem that made the project successful.
Instead, several relatively small questions were addressed at the appropriate stage:
None of these questions changed the fundamental product design.
They were manufacturing and project-management questions that appeared as the product moved from a tested prototype toward repeatable production.
That distinction is important.
A product can already be technically validated and still require additional manufacturing decisions before it is ready for stable production.
In this case, the goal was not to redesign the rally data logger.
It was to turn an already validated product into a manufacturing process that could be repeated with controlled sourcing, assembly, inspection, testing, and production risk.
That is often where the transition from prototype to production really happens.
A smooth OEM manufacturing project is rarely the result of one major decision.
More often, it comes from resolving a series of smaller questions before they become production problems.
The BOM needs to remain under control.
Component sourcing needs to be realistic.
Manufacturing capabilities need to match the actual product.
Testing needs to reflect production conditions.
And production processes need to accurately execute the customer’s approved requirements.
For industrial and performance-critical electronics, these details can have a much larger impact than the difference between two suppliers’ initial quotations.
The role of an experienced EMS partner is therefore not simply to build the product after receiving the files.
It is to understand the manufacturing requirements behind those files and identify the practical issues that matter before the product enters mass production.
For this UK rally data logger project, that meant moving from a tested prototype to production without introducing unnecessary changes to the product, while making the manufacturing process more controlled and practical.
A validated product is not automatically a repeatable production process. The transition happens through the manufacturing decisions made in between.
