Key Takeaways
- DPM readability depends on contrast, and contrast depends on lighting. Adjust your lighting method and angle before you change your camera or lens.
- Incremental read rate improvements can be transformative. Increasing first-pass reads by a few percent can significantly reduce scrap, rework, and costs.
- Barcode verification is an important factor; it helps determine whether a code will be readable across devices. Skipping verification introduces unnecessary risk.
Why Can’t a Laser-Based Scanner Read a DPM?
Your label reader fails to read a DPM because it searches for color contrast that does not exist. Black and white labels have high contrast. A laser-etched Data Matrix code, for example, has no contrast; the mark and the background are the same color. What remains is surface contrast: the physical geometry of the etched cells scatters light differently than the smooth background. If your reader’s lighting geometry doesn’t exploit that scattering effect, the camera sees a uniform gray rectangle and reports “No-Read.”
Only image-based barcode scanners and machine vision systems can read DPMs.
This challenge attracted industry attention early. As far back as 2007, Control Engineering reported that direct part marking methods create “little or no contrast between the ‘marked’ elements of a symbol and the background,” and that a new global guideline was created to “fill perceived gaps between existing print quality standards and the direct mark environment”. Two decades later, that gap remains the root cause of most DPM reading failures on factory floors.
The solution starts with lighting geometry. DPM reading requires illuminating the part at an angle that converts surface topography into a high-contrast image the decoder can process.
Dark field illumination is best for dot-peen marks, laser-ablated codes, or any raised surface topography.
Diffuse dome lighting works well for highly reflective, curved, or textured surfaces.
Lighting Techniques and their DPM Reading Effectiveness
| Lighting Technique | How It Works | Best For |
|---|---|---|
| Dark Field Illumination | Low-angle light skims the surface. Raised features (dot peen bumps, laser ridges) scatter light into the camera and appear bright. The smooth background reflects light away and appears dark. | Dot-peen marks, laser-ablated codes, any raised surface topography |
| Bright Field Illumination | Coaxial light hits the part straight-on. Rough, light-scattering marked areas appear dark. The polished, mirror-like background reflects light directly back into the camera and appears bright. | Laser-etched codes on polished or machined metal, marks darker than the substrate |
| Diffuse Dome Illumination | Light bounces inside a hemispherical dome coated with a highly reflective, matte surface, then cascades onto the part from all directions simultaneously. This eliminates hard shadows and mutes specular reflections. | Highly reflective, curved, or textured surfaces where single-direction lighting creates hotspots or uneven coverage |
| Structured Light | A known pattern (grid, stripes, or dots) is projected onto the part. The distortion of that pattern reveals the 3D topography, allowing the system to map dot peen indentations or laser engraving depth independent of surface color. | Low-contrast marks on bead-blasted, cast, or uniformly scattering surfaces; applications where pure reflectance fails |
| Multi-Angle / Multi-Shot Illumination | The system captures several images in rapid succession, each under a different illumination segment (e.g., dark field north, south, east, west, and bright field center). Software composites these into a single, glare-free image or attempts decoding on each individually. | Parts with unpredictable reflectivity, mixed surface finishes, or where the optimal lighting direction cannot be fixed—common on complex automotive and medical components |
| Polarized Illumination | A polarizing filter on the light source aligns the light waves, and a cross-polarized filter on the camera blocks the waves that specularly reflect off the shiny surface. Only the light scattered by the mark itself reaches the sensor. | Laser marks on highly polished, mirror-like metals where direct glare would otherwise saturate the entire image |
Is One Type of Marking Method Easier to Read than the Others?
No method of direct-part marking is inherently easier to read than others. A shallow dot-peen mark on cast iron produces irregular, low-contrast depressions that blend into the naturally rough surface texture. A poorly optimized laser mark on stainless steel creates micro-cracking and a barely perceptible color shift that vanishes under the wrong lighting angle. Both scenarios produce unreadable codes.
ETO Magnetic, an automotive supplier producing millions of parts annually, found their dot-peened DPMs suffered from thermal and mechanical distortion during production. Tool wear created differences in impression depth and angle. “When the last generation of code readers attempted to read these dot peen codes, up to four percent of parts were determined to have unreadable codes,” said Process Planner Klaus Schwanz. “This rate was clearly too high for us. In our manufacturing processes, a product with an unreadable code is treated as a reject.”
ETO Magnetic’s breakthrough came from treating the reading challenge holistically. Their original solution tried to read dot-peened DPMs the same way it read label codes, but the reflection from marking method made it ineffective. The new Cognex solution consisted of In-Sight vision systems, integrated lighting, and flat white lighting, which made dot-peen markings easily recognizable.
With the Cognex solution, ETO nearly eliminated instances of unreadable codes, achieved end-to-end traceability, and detected and processed material faults.
While all methods of direct-part marking can produce readable codes, a single lighting configuration and vision system cannot reliably read every type of DPM.
Expert Insight: What are Some Lighting Factors Nobody Considers when Reading DPMs?
Daniel Lapidus, Senior Applications Engineer, “One customer was assembling different parts, each of which had a DPM produced using a different marking method. One used low-angle dot peening to create DPMs, another used high-angle dot peening, while the third used laser etching.
The assembler had no chance at reading all three codes consistently, the lighting configuration for one did not work for the other two. Upstream manufacturers had to retool lines to be on the same marking methodology.
The inability to read DPMs, whether through normal wear-and-tear which obstructs the code or a mismatch of lighting and reading technology, can shutdown entire operations.”
What Does it Cost when a DPM Doesn’t Read?
In high-speed operations, a few seconds can have an immense ripple effect and shutdown production. If a manufacturing line produces 1,800 parts per day, failing to read just 3% translates to 54 parts needing manual intervention. And that means one full-time operator is dedicating their entire workload to rework.
Each no-read event may only stop the line for seconds, but those seconds compound. Kia Motors saw this clearly when they shifted their production to a six-speed transmission line. Their previous barcode reader would frequently stop manufacturing equipment due to code reading failures. The manual operation couldn’t match the manufacturing cycle, and operators risked installing wrong parts when codes were misread.
Expert Insight: What Slowly Degrades DPM Readability that Nobody Monitors?
Daniel Lapidus, Senior Applications Engineer, “Another customer was reusing plates with DPMs for a part carrier system. Over time, the DPM’s readability rapidly degraded, they eventually went from a 99% read rate to 80%, so one in every five codes was unreadable. The company had to pull carrier systems off the line and send them to maintenance for rework.
Rather than grading the codes and verifying if they were readable, the customer tried to upgrade their code reading solutions to no avail.
Both of those situations ended up being very expensive, there was a ton of downtime, and had to upgrade and retool their systems.
Having a code verification system is very important, knowing your codes are readable goes a long way.”
Barcode verification grades code quality based on international standards to ensure they can be read anywhere in the supply chain.
Verification: Ensure Readability and Compliance
Reading decodes the data. Verification grades the physical mark against an objective standard. These are fundamentally different processes and confusing them introduces risk that compounds across a part’s entire lifecycle.
When you read a DPM, you answer one question: “Can my system extract the data right now?” Barcode verification answers two questions
- “Will this code adhere to different regulations and standards?”
- “Will any compliant scanner, anywhere in the supply chain, be able to read this mark reliably?”
ISO/IEC 29158, released in 2020, is the international standard purpose-built for this challenge. It’s based on the international standard for label-based codes, but accounts for adaptive contrast measurement on reflective surfaces, modified binarization for uneven backgrounds, and tilted illumination for evaluating surface-contour marks like dot peen.
The verification-ensures-readability principle extends beyond manufacturing. In regulated industries, unreadable or low-quality codes create “traceability gaps” that complicate recalls, quality audits, and regulatory submissions. A verified mark with a grade of B or higher predicts reliable scanning across diverse equipment and conditions. An unverified mark—even one that your reader decoded today—offers no such assurance.
How Do I Read DPMs on Curved, Polished, or Machined Surfaces?
Curved surfaces create uneven illumination and focal plane distortion. Polished surfaces generate specular reflections that saturate the sensor. Machined surfaces introduce background texture—milling marks, grinding patterns, tool paths—that the decoder must distinguish from the code elements.
The solution to overcoming these challenges consists of:
- Lighting geometry addresses surface curvature and reflectivity. Computational imaging captures several images from different light directions and composites them, effectively canceling out the specular hotspot that would blind a single-angle setup.
- AI-powered decoding algorithms handle low contrast and background noise. Traditional decoders apply a global threshold to convert the image to binary: every pixel above a certain brightness becomes white; every pixel below becomes black. On a machined surface, that approach fails because the background texture produces pixel values that straddle the threshold. AI-based algorithms learn to separate code structure from surface noise.
- Image formation technology and HDR+ eliminate the exposure trade-off that shiny or textured metals force upon a standard camera. HDR+ retains image detail in both the brightest and darkest regions so the vision system processes a clean, uniformly lit image.
Fixed-Mount, Cobot-Mounted, or Handheld: Which DPM Reader Architecture Fits Your Line?
The decision hinges on three variables: throughput speed, part variability, and access constraints.
Fixed-mount readers deliver the highest throughput and repeatability. Integrated into the line at a known position with controlled lighting, they achieve first-pass read rates above 99% when the part presentation is consistent. They excel in high-volume production where every part arrives at the same orientation. For traceability programs requiring Ethernet connectivity and direct PLC communication, fixed-mount barcode readers with industrial protocols eliminate middleware.
Cobot-mounted readers provide flexibility for high-mix, low-volume production. When parts vary in size, shape, or code location, a collaborative robot positions the reader dynamically. This architecture handles complex geometries—engine blocks with codes on multiple faces, for instance—at the cost of cycle time. Each repositioning adds seconds.
Handheld readers serve rework stations, quality audits, and applications where parts are too large to move. In Kia Motors’ implementation, handheld readers served as backup systems for offline verification and rework. For quality teams managing dock-to-stock inspections or field service verification, handheld barcode scanners provide the same decoding algorithms in a portable form factor.
| Deployment Architecture | Best Application | Trade-off |
|---|---|---|
| Fixed-Mount | High-volume, consistent part presentation | Least flexible; requires consistent line of sight to the code |
| Cobot-Mounted | High-mix, variable geometry parts | Slower cycle time; higher integration complexity |
| Handheld | Rework, audit, large/stationary parts | Operator-dependent; not suited for inline speed |
Fixed-mount readers excel in high-volume production.
Handheld readers are portable for offline stations.
Are Your DPM Codes Compliant with Industry Regulations?
Compliance with DPM regulations requires understanding both the marking standards and the verification standards that apply to your industry. Automotive manufacturers follow the Automotive Industry Action Group (AIAG) guidelines. Aerospace suppliers adhere to SAE AS 9132 and MIL-STD 130. Medical device manufacturers must comply with FDA Unique Device Identifier (UDI) requirements. Each framework demands traceability, but the specific requirements for mark durability, readability, and verification differ.
Understanding these regulations and how to maintain compliance across marking technologies, substrates, and supply chain handoffs is critical. Direct part marks and their quality are essential for tracking and traceability efforts as well as maintaining compliance with government and industry regulations.