Additive manufacturing quality monitoring

Companies working in this area include Velo3D Inc., EOS Electro Optical Systems GmbH, Nikon SLM Solutions AG and Interspectral. Velo3D combines sensing and software for build tracking, while EOS offers systems with in-process monitoring capabilities. Nikon SLM Solutions has patented a sensor-based monitoring simulation method, and Interspectral develops software for visualizing process data and detecting anomalies.

Interspectral, FABRX and CurifyLabs Oy are startups active in related monitoring and quality-control work. Interspectral develops process-data visualization and anomaly-detection software for metal additive manufacturing. FABRX works on pharmaceutical 3D printing, including non-destructive dose verification, while CurifyLabs Oy develops automated medicine-compounding workflows designed for validated production and quality control.

Oak Ridge National Laboratory, the U.S. National Institute of Standards and Technology and Carnegie Mellon University are active research organizations in this field. Their work includes optical monitoring, melt-pool analysis, spatter measurement and wire-arc additive manufacturing. Wuhan University also researches powder-bed defect detection and melt-pool imaging, while Johns Hopkins University works on digital calibration for part qualification.

The recorded figures list $88.98 million across 107 grants since 2020 and $1.24 billion across 22 funding rounds. These are aggregated funding records, not necessarily amounts dedicated only to quality monitoring. For example, some listed awards concern broad institutional or research activity, so the totals should be treated as an indication of recorded funding rather than a clean market-specific estimate.

General Electric Company leads the listed patent holders with 9 patents, followed by JEOL Ltd with 8. Interspectral has 3, while The Penn State Research Foundation and Siemens Energy Global GmbH & each have 2. These counts identify the leading holders in the available patent data; they do not indicate the strength, scope or commercial status of each patent.

No: the data lists 61 patents in the last 24 months and a patent-growth change of -20.6%. This contrasts with publication activity, which grew 8.9% in the last year, to 867 publications. The difference suggests research output is expanding while recent patent activity is lower, though patent counts alone do not show whether particular methods are nearing commercial deployment.

Recent developments include the University of Arizona’s RAMPSeg image-segmentation method, published in October 2026, and Interspectral and Pankl’s expanded quality-assurance collaboration in June 2026. Research is also exploring standards-based AI monitoring, online anomaly detection and layer-resolved measurement for binder jetting. These directions aim to make sensor interpretation more reliable and connect monitoring to production decisions.

The main approaches are visible-light imaging, thermal imaging, optical emissions sensing, acoustic sensing, data-driven analytics and post-build inspection. Imaging and emissions sensing capture optical evidence of process behavior, while thermal and acoustic methods add temperature or sound signals. Analytics combine measurements to identify patterns, and post-build inspection evaluates the finished part. In practice, combining methods can provide complementary evidence.

Applications include medical and dental manufacturing, automotive and motorsport, energy and power, industrial manufacturing, and aerospace and defense. Medical and dental work is the largest listed application area, with 656 documents; industrial manufacturing has 327, energy and power 274, automotive and motorsport 373, and aerospace and defense 64. Monitoring supports defect detection and process assurance across these uses.

The United States leads the listed activity with 6,026 documents, followed by Germany with 2,389 and the United Kingdom with 2,029. China has 1,929 documents, and Japan has 1,191. These figures show where documented activity is concentrated; they do not by themselves measure commercial adoption, research quality or the number of active monitoring suppliers.