Small Stamped Brackets: 32 Points In 10 Seconds

Overview

A small stamped bracket is a different inspection problem from a subframe, even though the characteristics are similar. On a large part the camera has to cover distance; on a small part it has to fit inside a cycle that is often under ten seconds, and the constraint moves from optics to handling.

A bracket measuring a few hundred millimetres fits comfortably in one field of view, so all of its features can be measured in a single exposure. What then dominates the cycle is putting the part in front of the camera, holding it still and taking it away, and that is where a small-part station earns or loses its place on a line.

A station measuring thirty-two points on small structural parts in ten seconds, covering sixty to eighty vehicle variants, is a useful example of the arrangement. Big Bird Industrial (formerly Shimada Big Bird Industrial) has been building cleaning and deburring systems for automotive powertrain components since 2001, and the figures here come from production inspection stations.

1. Why Small Parts Are Different

Three properties change when the part is small, and each one shifts the design effort somewhere else.

One View Covers Everything

A bracket in a single field of view does not need multiple cameras or a moving sensor, which removes the coordination problem that large-part stations spend most of their engineering on. The whole part becomes a two-dimensional measurement with a known scale.

The Cycle Is Mostly Handling

A ten-second cycle on a small part might be one second of exposure and nine seconds of loading, positioning and unloading. That inversion means improvements come from the handling concept rather than from a faster camera, which is the opposite of what a buyer expects.

2. The Thirty-Two Points

The point count is set by the drawing rather than by the sensor, and the number itself is less important than what each point is for.

What Gets Measured

On a stamped bracket the controlled characteristics are usually hole positions, hole diameters, edge distances and the positions of locating features relative to each other. Thirty-two points is a typical count for a bracket family, and it covers the features that affect assembly rather than every dimension on the drawing.

Why The Count Is Not The Performance

Measuring more points costs almost nothing in a single exposure, so a station quoting a high count is not necessarily more capable. What matters is the accuracy on the points that matter and the time it takes to get the part in and out.

3. Fixturing Without Slowing Down

A small part still has to be held, and the holding is where the cycle is lost.

Nest And Clamp Versus Drop And Measure

A precise nest with clamps produces a repeatable pose and takes time to load. A loose nest that presents the part and measures it in place is faster but requires the software to establish the pose from the part's own features. Small-part stations increasingly use the second approach, because the measurement can find the datum that the fixture used to provide.

Why Gravity Helps

A small flat part rests repeatably on a flat surface under its own weight, which is a more consistent datum than a clamp on a stamped edge. Presenting the part face down on a reference plate and measuring from above removes the clamping variable entirely.

4. Covering Sixty To Eighty Variants

A bracket family is where a small-part station justifies itself, because one cell replaces a set of gauges.

5. Where A Small-Part Station Fails

The failures on small parts are rarely optical, and they repeat in the same three places.

Part Presentation

A stamped bracket that arrives bent, burred or with a hanging tag is a part the station cannot measure. The station rejects it as unmeasurable rather than as out of tolerance, and that distinction matters because one is a process problem and the other is a quality problem.

Reflective Surfaces

Oiled or polished small parts produce highlights that disturb edge detection, and the effect is worse than on a large part because the whole field of view is close to the surface. Diffuse illumination and a shorter working distance with a longer focal length are the usual mitigations.

Variant Identification

With eighty variants available, a station that runs the wrong recipe produces a pass on the wrong part. Identifying the variant from the part rather than from the operator's selection is what prevents that, and it is worth the extra second.

6. Specifying One

A specification for a small-part station should state the feature list per variant, the accuracy required per feature, the cycle including handling, the variant count and the method for identifying a variant.

It should also state what happens to a part that cannot be measured. A station without a defined path for an unmeasurable part will either pass it or stop the line, and neither is a decision anyone made deliberately.

7. The Comparison That Decides It

A fixture is faster per part and needs no recipe. A vision station is slower per part and replaces a rack of fixtures with one cell.

The comparison therefore turns on variant count and volume: a bracket produced in hundreds of thousands per year justifies its own fixture, and a family of sixty variants produced in thousands each does not. The arithmetic is the same as it is for any inspection choice, and the cycle figure is only one of its inputs.

8. Throughput, Buffer And The Real Cycle

A ten-second inspection is only useful if the line can present a part every ten seconds, and that is rarely how a stamping line runs.

Where The Part Comes From

Brackets usually arrive in stillages or on a conveyor from a press line that runs at its own rhythm. A station that measures in ten seconds but receives a part every ninety seconds is idle most of the time, and the ten-second figure describes a capability rather than a throughput. Stating both figures is what allows a buyer to size the buffer rather than the station.

Measurement Accuracy In Numbers

The accuracy statement for this class of station is agreement with a coordinate measuring machine of about 0.1 mm on hole positions, with repeatability better than that, measured at the working distance actually used. A hole position measured to 0.1 mm against a tolerance of half a millimetre consumes a fifth of the band, which is workable; the same accuracy against a tolerance of 0.2 mm is not.

9. Where The Station Sits

A ten-second station placed at the end of a press line sees parts at the press rate, not at its own. Placed after a buffer, it sees parts at its own rate as long as the buffer holds.

Before Or After Cleaning

A dimensional station is usually placed before cleaning, because oil on the surface scatters light and changes apparent edges. Where the part is measured after cleaning, the measurement happens on a dry part, which is more stable optically but adds the risk that a handling mark from the transfer is measured as a defect.

Before Or After Assembly

Measuring before assembly catches a bracket that is out of tolerance at the point where it can still be sorted. Measuring after assembly catches the case where a bracket was fitted to the wrong location, which is a different question and usually needs a different station.

Both placements are legitimate. What matters is that the specification names one, because a station designed for the wrong position in the flow measures the right features at the wrong moment.

Frequently Asked Questions

Can A Small Part Station Run At Five Seconds?

Only with automated handling. At five seconds the exposure is negligible and the loading dominates, so the cycle is set by whatever presents the part.

Does Measuring More Points Slow The Cycle?

Not in a single exposure. The processing cost of additional points is small compared with the handling time, which is why the point count is rarely the constraint.

How Is A Variant Identified?

From a code if one is present and legible, or from the geometry of the part. Geometry is slower but removes the dependency on a marking.

What Accuracy Is Realistic?

The usual statement is agreement with a coordinate measuring machine of around a tenth of a millimetre on hole positions, with repeatability better than that. The figure should be tied to the working distance actually used.

Conclusion

A small stamped part fits in one field of view, so its inspection is limited by handling rather than by optics. That is why a ten-second cycle is mostly loading, and why the handling concept decides whether the station works.

The variant count is what justifies the station, and the variant identification is what protects it. A cell that replaces a rack of fixtures is valuable only if it runs the right recipe for every part it sees.