For industrial equipment manufacturers serving the oil and gas and energy infrastructure sectors, a macro-level event can eventually become a very practical manufacturing problem:
Customers need more equipment, and they need it faster than originally planned.
A valve actuator control device we manufacture for a customer is a good example.
The project had already been in stable repeat production for some time.
Under normal demand conditions, the customer used an established VMI arrangement to hold long-lead components in advance, allowing the product to maintain a regular delivery cycle of approximately two weeks.
Then demand changed sharply.
The customer updated its full-year forecast, increasing the expected annual purchasing volume by approximately 60% compared with the forecast provided at the beginning of the year.
It is important to clarify that 20,000 units were not the entire 60% increase. They were the first and highest-priority requirement within that additional annual demand.
The customer wanted those first 20,000 units delivered as quickly as possible, with a target of four weeks or less.
And this was not expected to be a one-off requirement.
The customer expected multiple subsequent POs of around 20,000 units each for the rest of the year, with a similar delivery target of approximately four weeks.
So this was no longer simply a matter of handling one urgent order.
The real question became:
How do you adjust a stable manufacturing system when the annual demand baseline suddenly increases by approximately 60%?
Facing a sudden increase in end-market demand, the customer was under direct pressure from its own customers as well.
They could have taken a much simpler approach:
20,000 units. Four weeks. Please deliver.
But that would not create additional material or production capacity.
The customer understood that.
So the first thing they did was update the full-year forecast.
They told us that expected annual demand had increased by approximately 60% compared with the forecast provided at the beginning of the year.
They then clearly identified the first 20,000 units within that additional demand as the highest-priority requirement.
They also gave us visibility of what was expected to follow:
multiple additional POs of around 20,000 units each, with a target delivery cycle of approximately four weeks.
That gave the supply chain something extremely valuable:
Visibility.
It did not remove the problem.
But it gave both sides time to prepare for a higher demand level before the full volume turned into a series of production orders.
The customer was also managing the pressure on its own side.
Rather than simply transferring the end-market pressure downstream, they worked with their own customers to create some delivery flexibility while the manufacturing ramp-up was taking place.
For a large OEM, that is a meaningful distinction.
To understand what happened next, one point has to be made very clear:
This was not normally a seven-to-eight-week product.
Under the normal demand level, the customer and Fumax were already using VMI to prepare long-lead materials in advance.
That supported a regular delivery cycle of approximately:
Two weeks.
The VMI inventory built at the beginning of the year was based on the original full-year forecast.
So the original production model was working as intended.
The problem was that the annual forecast suddenly increased by approximately 60%.
The existing VMI inventory obviously could not cover that additional volume.
The materials required for the incremental demand therefore had to be purchased again.
Several critical components had normal lead times of approximately six weeks.
As a result, the incremental demand could no longer follow the normal two-week delivery cycle.
Once new long-lead materials had to be purchased, the expected delivery cycle for the additional volume stretched to approximately:
seven to eight weeks.
So the factory had not suddenly become a two-week factory and then turned into an eight-week factory.
What changed was the material position.
The established two-week production model was still there. The additional demand was simply outside the VMI coverage that had already been built.
We reviewed the BOM and quickly identified several critical components.
They included:
The normal lead time for these components was around six weeks.
There was another important constraint:
There were no previously validated alternatives.
This product had already been in production for years.
A shorter delivery target did not justify replacing validated components with something that merely appeared similar on paper.
So the real question was:
How do we secure the additional material without changing the customer’s validated product definition?
The customer and Fumax then worked from different positions in the supply chain.
For one critical MCU, the customer used its position as the end customer to communicate directly with the original manufacturer and ultimately secured priority supply.
For another critical power-management component, the customer provided an LOA, allowing Fumax to support the sourcing process.
For other long-lead components that could not be covered through normal channels, such as MOSFETs, we accelerated procurement through distribution.
This meant accepting a higher purchase price for some components in exchange for shorter availability.
It was not the cheapest sourcing solution.
But at this stage, the priority was clear:
Remove the material bottleneck created by the additional demand.
At the same time, we recognized that simply finding enough material to start production was not enough.
The real goal was to establish a supply position capable of supporting the incremental demand over the subsequent production cycles.
Otherwise, the bottleneck would simply move from purchasing to production.
After reviewing the material and production situation, the reality was clear.
The customer wanted the first 20,000 units within the incremental demand delivered as quickly as possible, with a target of four weeks or less.
But newly purchased long-lead materials could not arrive instantly.
So the four-week figure was a customer target to work toward, not a number that could simply be guaranteed under the existing material position.
We therefore used the existing VMI inventory as the first available production resource.
During the initial four-week window, we consumed that existing inventory and delivered approximately:
7,000 units.
The full first 20,000-unit requirement was not completed within that initial four-week window.
At the same time, however, several other activities were progressing in parallel:
New materials were being expedited.
New production equipment was being purchased.
Additional operators were being recruited and trained.
Custom-part suppliers were being pushed for priority production.
Testing resources, fixtures, stencils, and special consumables were being prepared.
So the first four weeks were not simply a delivery window.
They were also the critical window for building the manufacturing capacity required by the new demand level.
Even once the material constraints started to ease, another issue became clear.
The existing production capacity had been configured around the original demand level.
But the customer had already told us that the higher demand would not end with the first 20,000 units.
There would be multiple subsequent POs of approximately 20,000 units each, with a target delivery cycle of around four weeks.
Simply adding overtime, inserting rush orders, or taking capacity away from other customers might solve one shipment.
It would not create a sustainable production model.
What we needed was:
new production capacity.
That is when the project moved from an urgent order response into a broader capacity ramp-up.
We reviewed the full manufacturing flow:
Material → SMT → post-soldering → testing → custom parts → final assembly → packaging
And then we started addressing several constraints in parallel.
The SMT team recalculated the equipment required to support the new order level.
The question was not:
How much more can we squeeze out of the existing lines?
It was:
What additional capacity is required if approximately 20,000 units continue to enter the system and the target delivery cycle remains around four weeks?
Based on that evaluation, new equipment was purchased immediately.
From procurement and delivery through installation, commissioning, and production launch, the process took approximately:
three weeks.
The purpose was not to temporarily redirect the existing factory capacity toward one customer.
It was to create additional manufacturing capacity.
The post-soldering and manual assembly teams recalculated the additional labor required for the increased volume.
HR launched an urgent recruitment process.
At the same time, the production team selected two experienced employees to become new team leaders for the additional operators.
Because some long-lead materials were still being expedited, that waiting period was used for training.
The new operators were trained against the existing SOPs and familiarized with the product and assembly process before the full material flow was available.
So the ramp-up did not happen in a strict sequence.
Equipment was being added.
People were being recruited and trained.
Materials were being expedited.
These activities were happening in parallel.
Electronic components were only one part of the product.
The finished unit also depended on customized plastic housings and PVC membranes.
If the PCBA was ready but the housing was not, the finished product still could not ship.
So purchasing simultaneously went back to the custom-part suppliers and renegotiated production schedules.
The objective was straightforward:
Secure production priority.
To encourage suppliers to move their production schedules forward, we accepted certain commercial adjustments, such as higher advance-payment ratios and other conditions where appropriate.
When the overall delivery window is compressed, a supplier’s normal production queue becomes part of your own delivery lead time.
That means capacity ramp-up does not happen only inside the EMS factory.
It also happens throughout the surrounding supplier network.
At the same time, the warehouse team worked closely with purchasing to track the complete production preparation status.
The requirement was not limited to:
MCUs, MOSFETs, connectors, and other BOM components.
It also included:
testing fixtures, production jigs, stencils, solder paste, red glue, and other production consumables.
This particular project also used some special consumables that were different from the materials used in the factory’s regular production.
That meant they required separate preparation and tracking.
These items may not look like major materials.
But when the delivery target becomes four weeks, a missing fixture or a special consumable that has not arrived can become a bottleneck just as effectively as a missing IC.
So the production ramp-up covered the entire manufacturing chain, not just the BOM.
This is one of the most important parts of the case to describe accurately.
The customer target was:
The first 20,000 units within the incremental demand, delivered as quickly as possible, with a target of four weeks or less.
The actual result during the initial four-week window was:
Approximately 7,000 units delivered.
Those units were produced primarily by using the existing VMI inventory.
At the same time:
Critical materials continued to be expedited.
New SMT equipment was purchased, installed, and commissioned.
Additional operators were recruited and trained.
Custom-part suppliers adjusted their production priorities.
Testing resources, fixtures, stencils, and special consumables were prepared.
So the first four weeks were not simply a failed attempt to complete 20,000 units.
They were the period in which the existing inventory was converted into deliveries while the manufacturing system was being rebuilt for the higher demand level.
Some of the customized plastic housings did not arrive fully according to the original schedule, resulting in approximately one additional week of delay.
But this did not completely block production.
The reason was simple:
The PCBA could be manufactured first.
The electronic assembly could continue moving forward while the plastic parts were still being completed, with final assembly taking place once the housings became available.
This allowed the upstream manufacturing process and downstream assembly process to be separated where practical.
A delay at one downstream supplier therefore did not have to stop the entire manufacturing system.
For a compressed delivery schedule, that ability to separate production stages can itself become part of the manufacturing response.
This was the most important change in the project.
The first 20,000 units were simply the highest-priority first requirement within the approximately 60% increase in annual demand.
They were not the entire incremental demand.
The customer had already indicated that multiple subsequent POs of around 20,000 units each were expected for the rest of the year.
So the new equipment and additional personnel were not installed or recruited for one temporary PO.
They were intended to support:
a higher annual demand level.
As materials, equipment, people, and suppliers gradually came into position, the expanded production organization reached a capacity target of approximately:
10,000 units per week.
This was the capacity established to support the continuing incremental demand, rather than a temporary output level created simply to finish the first 20,000-unit requirement.
The customer’s ongoing target remained approximately four weeks per PO, while the expanded production capability provided a stronger capacity base for that continued order pattern.
In other words:
20,000 units represented the first high-priority requirement.
10,000 units per week represented the new ongoing production capacity established for the higher annual demand level.
The two numbers describe two different things.
The simplest way to give this project more capacity would have been to take existing SMT, post-soldering, and assembly resources away from other customers.
That could improve one customer’s priority temporarily.
It could also create delays somewhere else.
That was not the approach.
The objective was:
Increase the system’s actual capacity instead of simply redistributing the existing capacity.
New SMT equipment.
Additional operators.
New team leaders.
Training.
Supplier coordination.
Additional testing resources.
Additional production consumables.
This allowed the increased demand to be absorbed while the production of other existing orders remained stable.
For an EMS serving multiple industrial OEMs, that distinction matters.
True capacity expansion should mean:
the system can process more business without simply transferring the problem to another customer.
Looking back, the most interesting part was not that one department solved one problem.
Several constraints appeared at the same time:
Insufficient VMI inventory.
New long-lead material purchases.
Additional SMT capacity.
Additional labor.
Training for new operators.
Priority production for custom parts.
Additional testing resources.
Special production consumables.
Any one of these could have become the next bottleneck.
So the approximately 60% increase in the annual forecast did not simply test whether the factory could “make more.”
It tested whether the factory could:
Expand multiple parts of the manufacturing system at the same time without losing control of the system that was already running.
That is very different from simply working overtime to complete an urgent order.
Looking back, one of the most valuable parts of the customer’s approach was not the first 20K requirement.
It was the fact that they first gave us visibility of the approximately 60% increase in annual demand.
They were under direct pressure from their own customers.
But they did not simply turn that pressure into a series of:
“ASAP” POs
Instead, they first made the change in the annual demand baseline visible to the supply chain.
Then they identified the first 20,000 units as the highest-priority requirement.
They told us that more approximately 20,000-unit POs would follow.
And they worked on their own side to create some delivery flexibility while the manufacturing ramp-up was taking place.
That gave both sides time to prepare:
material, equipment, people, custom parts, fixtures, testing resources, and production organization.
A forecast cannot shorten a six-week component lead time.
But a timely forecast can allow the supply chain to see that six-week problem earlier.
And seeing the problem earlier gives you a chance to start solving it before the demand becomes a series of hard delivery commitments.
This project originally operated as a stable manufacturing system.
Under normal demand conditions:
VMI → approximately two-week delivery.
Then market demand changed.
The customer updated its full-year forecast by approximately:
+60%.
The existing VMI inventory was no longer sufficient to cover the incremental demand.
New material purchases pushed the expected delivery cycle for that incremental volume to approximately:
7–8 weeks.
The customer’s target for the first high-priority 20,000 units was to bring that delivery cycle as close as possible to:
four weeks.
During the initial four-week window, approximately:
7,000 units were delivered.
At the same time, the manufacturing system was being expanded through:
material acceleration, original-manufacturer coordination, distributor sourcing, SMT equipment expansion, recruitment and training, custom-part priority production, testing resources, fixtures, and special consumables.
As these resources came into place, a new ongoing production capacity of approximately:
10,000 units per week
was established to support the higher annual demand level and the customer’s subsequent approximately 20,000-unit POs with an approximately four-week delivery target.
And the expansion was carried out without disrupting the production of other existing orders.
That is what makes this project worth recording.
Stable production proves repeatability.
A sudden demand increase tests flexibility.
For an industrial OEM, both matter.
A mature EMS should not only be able to manufacture a product consistently under normal conditions.
When the market changes suddenly, it should also be able to identify the real bottlenecks, coordinate with the customer and the supply chain, add equipment and people, and integrate that new capacity into the manufacturing system already running.
The challenge is not simply to produce more.
It is to increase production capacity without losing control of the manufacturing system that is already running.
