A finished car seat gives away almost nothing about how it was built. The fabric hides the shell, the shell hides the frame, and the frame hides the decisions that actually determine whether the seat protects a child. So here is the whole process, stage by stage, from our own floor.
Ask most people how a car seat is made and they picture an assembly line. That is the last quarter of the story. By the time a shell reaches assembly, its safety performance is already fixed — decided weeks earlier by a mold designer and a materials choice.
This guide walks the full sequence. At each stage we have added what an importer should verify, because the same steps that build the seat are the steps where quality is either protected or quietly removed.
Every seat starts as a decision about who it is for. Height range, market, price point, install method. A seat for a 40 to 150 cm range needs a rotating base and a five-point harness. A booster for 125 cm and up needs neither. Those choices lock in before anyone draws anything.
For an OEM project, this is where the customer's drawings, reference sample, or brief arrives. For our own models, it is where the product team defines the target and the market it must be certified for. If you are unsure which type fits your market, our guide to car seat types maps the whole range.
Engineers turn the brief into a real structure. This stage sets the load paths — where crash force will travel, and what will carry it. The anti-rotation method is decided here too: a support leg braced to the floor, or a top tether strapped to an anchor. That single choice determines which cars the finished seat can be used in, which is why we cover it separately in support leg vs top tether.
Material grade is also fixed here. We use 100% virgin plastic. Recycled material is cheaper and behaves less predictably under impact, which is exactly the wrong trade in a safety product.
This is the stage almost nobody asks about, and the one that matters most.
A mold is a machined steel tool. Once it exists, every single shell that comes out of it inherits its geometry — its wall thickness, its rib layout, its tolerances. A well-cut mold produces 200,000 consistent shells. A poorly cut one produces 200,000 shells with the same weak spot.
Having an in-house mold shop is not about saving money. It is about speed of correction. When a test shot comes out with a thin wall in one corner, we walk to the mold shop and adjust it. A factory that outsources tooling has to describe the problem, ship the tool, wait, and hope.
That difference shows up most in OEM work. A custom shell almost always needs two or three tooling revisions before it is right. In-house, that is days. Outsourced, it is weeks each time.
Now plastic becomes structure. We run two methods, and a serious range uses both.
| Method | Produces | Used for |
|---|---|---|
| Injection molding | Solid, dense, precise parts | Bases, belt guides, ISOFIX housings, booster shells |
| Blow molding | Hollow, double-walled shells | Large one-piece shells that absorb and spread crash energy |
Neither is better. They solve different problems, which is why our flagship BW25 uses both in one seat — a blow-molded body over an injection-molded structure. The full technical comparison is in our molding explainer.
A shell alone is not a car seat. Several components go in before the cover does, and each has a job in a crash.
Everything that protects a child in a crash is hidden by the time the seat reaches a shelf. That is exactly why paperwork and testing matter more than appearance.
Covers are designed, cut, and sewn in our own sewing workshop. The fabric has to do more than look right. It must be non-toxic against a child's skin, breathable enough for long trips, colorfast so it does not fade unevenly, and durable through repeated washing.
This is also where most OEM personality lives. Custom colors, logo embroidery, and label design all happen here — and none of it requires new tooling, which is why appearance customization is affordable while structural change is not.
Shell, foam, frame, harness, buckle, hardware, and cover come together as one product. Assembly is repetitive work, and repetitive work is where small errors multiply — a harness routed slightly wrong, a connector not fully seated, a label applied to the wrong model.
Good assembly is therefore mostly about sequence and checking. Each seat is built in a fixed order, and functional parts are tested as they go in rather than only at the end.
Two different things happen here, and buyers often confuse them.
Type approval testing proves the design is safe. It is done once, by a recognized authority, and produces the certificate. Ongoing production testing proves that what leaves the line today still matches that approved design. Both are needed. A factory with only the first is showing you a photograph of its best day.
| Test | What it checks |
|---|---|
| Structural strength | Interface durability, five-point harness load-bearing |
| Impact | Frontal, side and rollover |
| Material | Flammability, toxicity, ROHS, REACH |
| Service life | Rotation and disassembly cycles of 10,000+ |
| Fabric | Color fastness, washing durability |
On top of third-party approval, we run COP (Conformity of Production) checks and crash-test 1 in every 5,000 seats we build. Our approvals are published as downloadable certificates: BW25, BW15, and BW26A R129/04. What each standard requires is covered in our R129 guide.
The last stage is more commercially important than it looks. Car seats are bulky, so a large share of what you pay to ship is air. How the product is shaped and boxed decides how many units fit a container — and that moves your cost per seat more than most buyers expect.
In our own loading, a compact backless booster fits roughly 2,190 units in a 40HQ, while a bulky high-back booster fits around 518. Same container, more than four times the units. We break the math down in the pricing and lead time guide.
If you take one thing from the nine stages, take this: the visible stages are the least decisive ones. Assembly and packing are execution. What separates a good seat from a dangerous one happens at mold making, material choice, and testing — three stages that leave no trace on the finished product.
That is inconvenient for buyers, because it means you cannot evaluate a car seat by handling it. You have to evaluate the factory. Practically, that means four questions:
Our supplier screening guide covers the rest of that process. To see the range these stages produce, browse the full product lineup, or tell us your market and we will walk you through the build.
In nine stages: sample confirmation, product development, mold manufacturing, molding the shell, fabric cover production, assembly, quality inspection, bulk production approval, and packing. Mold making and testing are the stages that most affect safety.
Usually a high-strength plastic such as polypropylene, chosen for its balance of strength, flex, and low weight. Quality factories use 100% virgin material rather than recycled, because recycled plastic behaves less predictably under impact.
A prototype typically takes about 7 to 15 days, and mass production about 30 to 40 working days after the deposit and confirmed details. New tooling adds time on top of that.
Behind the fabric there is a molded shell, energy-absorbing foam, often a built-in steel frame, a five-point harness with buckle, and anchoring hardware such as ISOFIX connectors with a support leg or top tether.
Ask whether they cut their own molds and mold their own shells, request a live video call from the molding line, and check that the safety certificate is issued in their own company name.
Conformity of Production testing is ongoing verification that seats coming off the line still match the approved design. Type approval proves one sample passed; COP proves production keeps passing.
Timelines and quantities shown are typical ranges and vary by model, volume and specification. Confirm all terms in writing with your supplier.