ISO/IEC 29158:2025, titled "Information technology, Automatic identification and data capture techniques, Direct Part Mark (DPM) Quality Guideline", was published in March 2025 and replaces ISO/IEC 29158:2020, which had been in force from December 2020 until its withdrawal in 2025 [S3][S9].
It is maintained by ISO/IEC JTC 1/SC 31 (Automatic Identification and Data Capture), ICS 35.040.50, and the 2020 edition ran to 33 pages, expanding the original 18-page TR 29158:2011 first issued in October 2011 [S1][S3]. The standard tells verifier vendors and application specifiers how to grade 2D marks whose reading device is a two-dimensional imager, the dominant configuration on automotive, aerospace, medical-device and electronics lines [S1][S3].
What ISO 29158 Actually Measures on a Direct Part Mark
ISO/IEC 29158:2020 defines alternative illumination conditions, new terms and parameters, and modifications to the measurement and grading flow defined in ISO/IEC 15415, with the result reported as a DPM grade instead of a 15415 grade [S3]. Cognex's published guidance confirms the 2020 update introduced changed grading scales and a new algorithm to better reflect how DPM marks are read in the real world [S5].
Four illumination geometries are normatively referenced: 30°, 45°, 90° (directional) and a hemispherical "dome" configuration, each chosen so the lighting casts shadows inside laser-ablated craters or dot-peen indentations to maximize signal-to-noise against the substrate texture [S4][S5]. ISO 29158 uses grading parameters such as Cell Contrast, Modulation, and Quiet Zone that differ from those of ISO 15415, enabling quality assessment of direct part marks on substrates including metal where grey-on-grey contrast can otherwise hinder readability [S4].
Per Euresys documentation, ISO 29158 is a modification and an extension of ISO 15415 designed for DPM codes, primarily Data Matrix and QR, with the symbology-specific syntax checked against ISO/IEC 16022 (Data Matrix) and ISO/IEC 18004 (QR) [S8].
Reading vs Verifying: Pass/Fail vs A-through-F Grading
A "readable" Data Matrix only proves that some scanner, including a consumer smartphone, could decode the data string at one moment, with no guarantee of robustness across the supply chain [S4]. A "verifiable" mark is one a calibrated verifier has graded against ISO 29158, with separate scores for each sub-parameter and an overall letter grade from A (4.0) down to F (0.0) on the standard's modified scale [S4][S5].
Cell Contrast (CC) measures the reflectance difference between marked and un-marked cells, the single biggest failure mode when a MOPA fiber laser produces too shallow an anneal on stainless or too dark a carbonization on plastics [S4]. Cell Modulation (CM) checks consistency across all cells, and is widely described in industry guidance as the most common grade-killer when laser power fluctuates or surface texture varies [S4]. Axial Non-Uniformity (ANU) confirms the symbol is geometrically square, while Fixed Pattern Damage (FPD) protects the L-shaped finder pattern and timing dots from quiet-zone encroachment [S4].
The combined grade is the lowest of CC, CM, ANU, FPD, plus the symbology-specific decode score, so improving one parameter does not lift the overall grade if any other remains in D or F territory [S4][S5].
Why ISO 15415 Fails on Metal and Plastic Substrates
ISO/IEC 15415 was written for high-contrast printed labels, black ink on white paper, and its 50% binary threshold and assumed fixed 90° illumination collapse when applied to grey-on-grey laser marks on steel, aluminum or titanium [S4]. The 29158 modification set, including four-angle lighting and a tuned threshold, is the engineering fix that lets the same 2D imager architecture produce stable, repeatable grades on a milled engine block or a printed circuit board [S1][S3][S4].
For procurement teams this matters on every part-marking spec sheet: if a supplier quotes a 15415 grade for a lasered Data Matrix on aluminum, the number is technically meaningless for process control and the mark has not really been verified to a DPM-appropriate method [S4][S8]. The right line item is an ISO 29158 (DPM) grade with the illumination angle explicitly called out, because the same code can grade differently at 30° versus dome depending on the mark depth and substrate reflectance [S4].
This grading discipline shows up directly in traceability loops on parts that downstream pass through pressure transmitter calibration cells, flow meter bodies and industrial valve housings, where a single failed decode on a heat-treated forging can scrap a serialized unit.
Edition History: 2011 TR, 2020 IS, 2025 Update
Edition 1 of the document, originally classified as a Technical Report, was ISO/IEC TR 29158:2011, published 2011-10 with 18 pages and withdrawn when the full International Standard was published [S1]. ISO/IEC 29158:2020 followed on 2020-12, expanding to 33 pages, and was also withdrawn in 2025 to make way for the current edition [S3].
ISO/IEC 29158:2025, published 2025-03-24, is described in the cataloguing entry as specifying quality test methods for direct part mark bar codes with tailored illumination and grading, aimed at verifier manufacturers and application specification developers [S9]. Because the 2020 publication stage sits at 95.99 (withdrawal) on the ISO life-cycle tracker, any spec written today should reference the 2025 edition, not the 2020 text [S3][S9].
The committee path is informative for buyers tracking change control: JTC 1/SC 31 took the 2020 revision from new project approval in November 2015 through three committee draft cycles, a DIS ballot in early 2020, and publication that December, a roughly five-year cadence that suggests the 2025 revision follows a similar but compressed review window [S3].
Choosing a Verifier and Lighting Setup for Production
For shop-floor use, the standard applies to any 2D imager, handheld or in-line, that is calibrated to the four-angle illumination set and can report the full DPM parameter list, not just a single overall grade [S1][S3][S8]. Cognex and Keyence publish product documentation explicitly referencing ISO/IEC TR 29158 / ISO/IEC 29158 verification workflows, and Euresys lists 29158 support inside Open eVision code-grading libraries, so the standard is multi-vendor rather than tied to one supplier [S5][S7][S8].
Substrate and marking process drive the right configuration: MOPA fiber lasers producing black-annealed marks on stainless, white-foam marks on plastics, or ablated marks on anodized aluminum each need different optimum angles, and a single 90° reading will under-grade marks that look clean at 30° or under dome lighting [S4]. Cross-hatching at roughly 0.05 mm spacing on 45°/135° angles plus a 4-module quiet zone around the symbol is the practical drafting rule repeatedly cited for moving a DPM mark from D to A on metal [S4].
On the process side, the same traceability loop is feeding a PLC on a marking station, so verifier output is normally pushed over PROFINET, EtherNet/IP or OPC UA into the same MES database that records torque on ball valve bodies and serial numbers on embedded part sub-assemblies.
Common Failure Modes and What to Fix First
If CC is the failing grade, switch MOPA pulse parameters to drive a true color-change anneal rather than a shallow oxide, and re-check the substrate finish, bead-blasted surfaces can physically lower contrast by scattering the verifier's illumination [S4]. If CM is failing, slow the marking head, stabilize part fixturing, and widen the wobble/overlap so each cell is filled by more laser passes, which averages out power fluctuations across the cell [S4].
ANU failures almost always trace back to motion: a conveyor that is indexing while the laser fires, a rotary axis on a turned part that is not synchronized, or a galvo lens operating at the edge of its field where f-theta distortion stretches cells. FPD and quiet-zone failures are the cheapest to fix: leave 4 modules of clean substrate around the symbol and never run a code to the edge of a chamfer or a bearing journal [S4].
For shop managers writing a PFMEA, the right failure-rate metric is "% of marks graded C or above at first read", not "% of marks decoded", because the first metric captures the 29158 quality distribution and the second hides the marks that are one vibration or one oil smear away from failing downstream [S4][S5].
Standards Lineage and What to Cite on Drawings
For procurement and quality, the canonical line on a drawing or CoC is: "Direct part mark shall be a Data Matrix (or QR) symbol verified to ISO/IEC 29158:2025, grade C or better overall, illumination per verifier manufacturer recommendation, with sub-parameter grades reported for Cell Contrast, Cell Modulation, Axial Non-Uniformity and Fixed Pattern Damage" [S3][S9]. That phrasing keeps the spec tied to the live edition rather than the withdrawn 2020 or 2011 texts [S1][S3].
The standard's lineage is consistent and traceable on the ISO site: 2011 TR, then 2020 IS, now 2025 IS, all under JTC 1/SC 31, all classified as DPM Quality Guideline, all in ICS 35.040.50 [S1][S3][S9]. Verifier manufacturers and application specifiers are the named audience in every edition, which means a purchasing team should expect a vendor's datasheet to state conformance to ISO/IEC 29158 and to publish the specific symbology specifications used in the test, not just a generic "DPM verified" claim [S1][S3].
Two signals worth watching in the next revision cycle: the diffusion of multi-imager dome verifiers that average out illumination angle sensitivity, and the integration of 29158 grades into digital-twin threads so each mark's sub-parameter history is stored alongside the digital twin for wind turbine blade production records that follow the same part through machining and assembly.