Use Case

Machine Vision for Semiconductor Molding Inspection

Semiconductor molding spans conventional packages, advanced packages, wafer-level and panel-level packaging. As molding processes, package structures and workpiece formats change, so do the imaging requirements, from compact in-machine inspection to process monitoring and large-area, high-resolution imaging.

Industrial camera integrated into automated semiconductor molding equipment for machine vision inspection of components on a production platform.
Machine vision system for automated inspection in semiconductor molding equipment.

Different molding processes create different vision requirements

Transfer and compression molding can be used across different semiconductor packaging applications. For machine vision, however, the inspection task is also shaped by the package structure, inspection area and process requirements.

Packaging type

Key process characteristics

Imaging requirements

Conventional packages

Package-level molding and encapsulation, commonly using transfer molding; compression molding is also used for selected package types

Localized inspection depending on the process step, with integration space, FOV and surface contrast as key considerations

Advanced packages

More components and increasingly complex package structures; transfer or compression molding can be used depending on the process

Fine-detail inspection across complex structures and materials, requiring suitable resolution, optics and illumination

FOWLP

Dies are molded into a reconstituted wafer, typically using compression molding

Die position and post-mold die shift measurement; wafer-scale inspection increases FOV and calibration requirements

FOPLP

Fan-out packaging extends the reconstituted structure from wafer to larger panel formats, typically using compression molding

Large-area inspection increases the challenge of balancing FOV, resolution, calibration, focus and throughput

semiconductor transfer molding process in three stages: load and clamp, resin transfer through the mold cavity, and cure and degate.
Three stages of semiconductor transfer molding: loading, resin transfer, and curing.
Diagram-showing-three-stages-of-semiconductor-compression-molding-resin-granules-loaded-onto-release-film-mold-closing-and-compression-under-vacuum-and-curing-and-release-of-the-molded-package.
Three stages of semiconductor compression molding: resin loading, compression, and curing.

Machine vision inside molding equipment

Machine vision can support multiple inspection points before and after molding. While transfer and compression molding have some inspection tasks in common, the imaging requirement can differ due to their resin-handling processes.

Diagram comparing a single-camera setup requiring a long working distance with a compact multi-camera configuration providing overlapping fields of view for semiconductor package inspection.
Comparison of single-camera and multi-camera configurations for semiconductor molding inspection, illustrating how multiple cameras enable wider coverage within limited machine space.

Integrating vision within limited machine space

For both molding methods, vision can be used around the molding process for tasks such as workpiece positioning, mold or process-area inspection, and post-mold inspection. The exact inspection point depends on the equipment design and package being processed.

As workpiece and molding areas become larger, using a larger sensor or higher-resolution camera to cover a wider FOV may seem like the straightforward approach. However, a larger FOV can require longer working distances or larger optics, which may not fit within the available machine space. In these cases, compact or distributed camera configurations, such as dart (27 mm × 27 mm) or ace 2 (29 mm × 29 mm) can provide greater integration flexibility while covering the required inspection area.

Discuss your requirement
Diagram showing a multi-camera inspection setup for semiconductor compression molding, comparing uniform resin compound distribution with uneven spread, gaps, and material pile-up.
Machine vision inspection of resin compound distribution in semiconductor compression molding, identifying uneven spread, gaps, and material accumulation before mold closing.

Resin dispense monitoring in compression molding

Compression molding introduces an additional imaging opportunity because the molding material can be dispensed or distributed directly into the mold before compression.

Vision can be used to monitor the resin dispensing area and determine whether the required distribution or pattern is achieved before the molding step. The inspection requirements depend on the resin type, dispensing method and process.

For these applications, the challenge is not only camera resolution. FOV, installation space and sufficient contrast between the resin and mold surface also need to be considered.

Advanced packaging scales the imaging challenge

In fan-out packaging, molding does more than encapsulate individual components. In mold-first FOWLP, dies are placed on a temporary carrier and over-molded to create a reconstituted wafer for subsequent processing. This introduces positional and dimensional challenges that are less prominent in conventional package-level molding.

Diagram illustrating die shift measurement on a reconstituted wafer after semiconductor molding, showing nominal versus measured die positions and X/Y offsets for downstream RDL alignment.
Machine vision measurement of die shift in wafer-level molding, comparing actual and nominal die positions to support downstream RDL alignment.

Measuring die shift after molding

During wafer over-molding, dies can move from their intended positions. As downstream interconnect dimensions become finer, knowing the actual die position after molding becomes increasingly important for subsequent processing.

Imaging can therefore move beyond defect detection toward die-position and die-shift measurement, providing coordinate information that can support downstream alignment and process control.

Diagram comparing a 300 mm wafer and a 600 × 600 mm panel, showing inspection field coverage, increased image acquisition requirements, and substrate warpage relative to the camera's depth of field.
Comparison of wafer-level and panel-level semiconductor molding inspection, highlighting increased field-of-view requirements, image acquisition demands, and warpage challenges for larger substrates.

Scaling from wafer to panel inspection

Moving from FOWLP to FOPLP extends similar packaging processes to much larger panel formats. The fundamental challenges—including die shift and warpage—remain, but become more difficult to manage as the processing area increases.

For optical inspection, a larger area must be covered while increasingly fine features still need to remain visible. This creates competing requirements for FOV, resolution and throughput. Panel warpage can further introduce height variation across the inspection area, affecting focus and measurement stability.

As a result, panel-scale imaging systems need to consider not only camera resolution, but also optics, focus strategy, calibration, acquisition architecture and data processing.

Designing your next semiconductor molding system? Discuss with our expert


As semiconductor molding moves from package-level applications toward wafer- and panel-level processes, the imaging challenge changes significantly. Larger inspection areas, die-position requirements and machine integration constraints need to be considered together when defining the camera, optics and illumination.
Application Engineer, Basler Asia

Imaging considerations for semiconductor molding

The imaging architecture depends on the inspection target, workpiece size, material characteristics and available machine space. Key considerations include:

Application challenge

Imaging consideration

Limited machine space

Camera size, optics and working distance need to fit within the available installation envelope

Larger inspection area

Balance FOV and resolution; a single larger camera may not always fit the required optical geometry

Different material surfaces

Molding compound, silicon and reflective metal can require different illumination and optical conditions to achieve sufficient contrast

Wafer / panel warpage

Height variation can affect focus and measurement stability across the inspection area

Higher inspection throughput

Larger FOV and higher resolution increase image data, making acquisition and processing speed increasingly important

Matching the vision system to the molding application

Semiconductor molding presents different imaging challenges depending on the molding process, package architecture and inspection task. From localized inspection inside molding equipment to wafer- and panel-level inspection, vision systems must balance FOV, resolution, image contrast, integration space and throughput. Camera, optics, illumination and image processing should therefore be selected together according to the application requirements.

Designing your next semiconductor molding system? Discuss with our expert

Products for this solution

Looking to implement a comparable solution? These products will help you.

Ready to get started together?

Learn more about our products, services, and how we can help you with your next project.