A British motorsport client recently approached us about moving production of a rally 4G data logger from the UK to China.
The prototypes had already been tested in the UK, and the initial production requirement was 100 units.
What followed was less about changing the product and more about working through the practical manufacturing details before production began.
The client was a British company, and the product was a 4G data logger for rally cars.
The device connects to the vehicle’s OBD interface, collects vehicle data, and transmits the data over 4G together with GPS position information. The product was intended for use in rally cars, so the electronics had to work in a very different environment from a typical consumer IoT device.
The prototype had already been tested in the UK.
The next step was production in China.
The initial requirement was only 100 units, but the first email was unusually complete. The client provided the PCBA manufacturing files, BOM, product background, application information, quantity, and company website for reference.
The commercial background was also explained clearly.
That meant we could send the technical information to our purchasing team for evaluation almost immediately.
There was no need for several rounds of emails just to establish what the product was, what it was used for, or what the customer was looking for.
We prepared the quotation and sent it over.
A few days later, the client came back with one question:
“Your quotation is about 8% higher than others. Did you use alternatives?”
We had not.
Our answer was simply:
“No.”
The client was satisfied with that answer.
There was no request to replace components or reduce the quotation through alternative parts.
We had also included reference pricing for 1,000 and 5,000 units, even though the immediate requirement was only 100 units.
That gave the client some visibility into how the pricing would change if the project moved beyond the initial batch.
With the quotation settled, the next discussion was about the factory itself.
The client did not have time to travel to China before placing the initial order, so we arranged a video meeting instead.
He had several specific questions prepared.
The first was about the mechanical side of the product.
The client wanted to know whether we could manufacture metal enclosures, what types of products we had previously built, what protection levels we had experience with, and whether we had relevant certifications.
We showed him examples of industrial controller enclosures we had manufactured, including products designed for protection against water ingress.
Rather than discussing enclosure manufacturing only in general terms, we could show him actual products and their construction.
The discussion then moved to the PCB.
The product used multi-layer PCBs and BGA components.
The client asked whether we were familiar with via-in-pad and how BGA quality was controlled during volume production.
We discussed the PCB assembly process and then moved into something that is easy to overlook when discussing BGA assembly: solder paste.
Selecting the appropriate solder paste is only one part of the process.
Storage, preconditioning, handling, and the temperature and humidity conditions in the SMT workshop can also affect the resulting solder joints.
We showed the client X-ray images from a previous project, including examples of acceptable and problematic BGA solder joints.
The images made the discussion more practical. Instead of simply saying that BGA quality was inspected by X-ray, we could show what the inspection actually looked like.
The client then moved to another component that was important to his design.
The product used a Quectel 4G module.
The client wanted to know how familiar we were with Quectel and, more importantly, how the module would be sourced for a relatively small initial production quantity.
We explained that we already had several volume-production projects using Quectel modules and established sourcing channels for them.
For an initial order of 100 units, there was no reason to force the customer to purchase thousands of modules simply to obtain genuine components.
We could also use authorized distribution channels that could supply smaller quantities.
The quotation included a detailed component cost breakdown, so the client could see the purchasing cost of the main components.
If the client already had a better sourcing channel, we could use his source instead.
There was also another option. If the client had access to a better price directly through his own supplier relationship, he could authorize us to purchase through that channel. This could allow the better purchasing terms to be used for the production order without changing the manufacturing arrangement.
With the enclosure, PCB assembly, and component sourcing questions addressed, the discussion moved to production testing.
The client’s original testing approach was based on the actual operating environment of the product.
He planned to use the real customer firmware and simulate the vehicle environment using an ECU.
The approach was understandable for prototype or engineering verification.
There was, however, a practical issue for production.
A suitable ECU for the test setup would cost approximately USD 200.
More importantly, the client did not have rework capability on his side.
So we asked a simple question:
“If a unit fails after shipment, do you have rework capability on your side?”
The answer was no.
That made the production test arrangement more important.
If a defective unit escaped production and reached the customer, there would be no local repair capability available. The unit would have to come back for investigation and repair.
We therefore discussed a different testing arrangement.
Instead of using the real operating environment for every production test, we proposed using two firmware versions.
The first would be specifically prepared for production testing.
Every unit would first be programmed with the production test firmware and tested through the production fixture.
This allowed the production test to focus on verifying the functions that needed to be checked on every unit without requiring the expensive ECU setup for every test cycle.
After passing the production test, the actual customer firmware would be programmed onto the unit.
We would then perform sample checks using the real firmware and the actual operating conditions.
During the video meeting, we also showed the client the basic working principle of the test fixture so that the proposed sequence was clear.
He agreed with the approach.
The final arrangement was therefore:
Production test firmware → 100% production testing → actual customer firmware → sample verification
This was more practical for production while still retaining verification with the actual customer firmware.
The project started with an initial quantity of only 100 units, but several practical manufacturing details had been discussed before the first units were built.
The enclosure, PCB assembly, BGA process, solder paste handling, Quectel sourcing, firmware programming, and production testing were all part of the conversation.
The first production run went smoothly, and the project subsequently moved into the normal production process.
The product had already been developed and tested in the UK. Our role was to take the existing product requirements, manufacturing information, and test requirements, and make them work together on the production floor in China.
The transfer was not about changing the product. It was about making sure the existing product could be produced consistently.
