Electron Beam Melting AI Patents: Leaders & White Space 2026
- Filing peaked in 2021 at 13 families, then fell back to flat-to-declining levels through 2022 and beyond — this looks like an early claiming wave, not a technology still accelerating.
- B33Y and B22F dominate the IPC mix, but G06T (image data processing) appears in 11 of 26 families, meaning most defect-prediction claims are built on vision and sensor-data pipelines, not generic ML.
- Europe and the US together receive 21 of the filings, with WIPO PCT filings still limited to 2 — most applicants have not yet gone for broad international coverage.
What this landscape covers
This dataset tracks patent families at the intersection of electron beam melting and electron beam powder bed fusion with machine learning process control, defect prediction, and parameter optimization. It spans filings from 2017 through the 2026 cut-off, drawing on 26 published families indexed across Europe, the United States, WIPO, China and Germany.
Because publication lags filing by roughly 18 months, the 2025 and 2026 figures in any trend line understate actual filing activity for those years. The IPC composition below is read at the family level, which is the fairer unit than raw document counts since it removes the effect of continuations and multi-jurisdiction refiling.
Filing trend and technology composition
The filing curve and the IPC mix together describe a field that filed hard around one peak year and has not sustained that pace since — while still splitting claims across additive manufacturing, powder metallurgy and image-processing classes.
A single peak year, not a steady climb
Filings rose from a single 2017 family to a peak of 13 in 2021, then dropped back toward the 2022 midpoint of 2 — consistent with an initial wave of process-control and defect-prediction filings rather than a technology still building momentum. The most recent years should be read with the publication lag in mind.
Vision and control classes sit alongside the AM core
B33Y (13) and B22F (12) confirm this is squarely an additive-manufacturing and powder-bed dataset, but G06T (11) shows that image data processing underpins most of the defect-prediction claims. G05B and G06F each appear in 5 families, and G06N — AI-model computing specifically — appears in only 2, suggesting most claims frame the invention around the manufacturing process rather than the underlying model architecture.
Shares are the percentage of the 26 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Electron Beam Melting AI and Machine Learning with Eureka
This page is one run against one query. Ask Eureka your own question about electron beam melting ai and machine learning and every answer comes back with the patent numbers behind it.
Try EurekaRepresentative and most-cited filings
Topology optimization for additive manufacturing with integrated defect prediction model
A controller with processing circuitry breaks a part design into elements, assigns an initial density to each, and determines the likelihood of a defect in each element using received manufacturing parameters and the assigned density. It evaluates defect volume across the elements and updates the part design or the manufacturing parameters in response, tying topology optimization directly to a defect-prediction model rather than treating the two as separate steps.Filed by RTX Corporation, published 2026-05-28.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | WO2023111542A1 | Defect identification in additive manufacturing based on time series in-process parameter data | 13 |
| 2 | US20230302539A1 | Tool for scan path visualization and defect distribution prediction | 12 |
| 3 | CN108446414A | 一种由3D打印多孔结构随机缺陷的逆向预测方法 | 7 |
| 4 | EP4249153A1 | Tool for scan path visualization and defect distribution prediction | 5 |
| 5 | US20230260103A1 | Computer-implemented, adapted anomaly detection method for powder-bed-based additive manufacturing | 5 |
| 6 | US11531920B2 | System and process for verifying powder bed fusion additive manufacturing operation as being defect free | 5 |
| 7 | US20210334685A1 | System and process for verifying powder bed fusion additive manufacturing operation as being defect free | 3 |
| 8 | EP4299290A1 | Method for developing additive manufacturing technology and 3D additive manufacturing system | 2 |
| 9 | EP3940630A1 | Computer-implemented, adapted anomaly detection method for powder-bed-based additive manufacturing | 2 |
| 10 | EP4467265A1 | Laser-plume interaction for predictive defect model for multi-laser powder bed fusion additive manufacturing | 1 |
Citation counts are drawn from a searched corpus and favour older filings; treat them as a signal of influence rather than of current technical importance.
Patent titles are shown in the language they were filed in, not translated, so that each record stays verifiable against the original filing — a translated title will not match in Eureka or in any national register. Each row carries its publication number; clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →What the numbers mean for a filing decision
Three patterns stand out once the filing trend, IPC mix and citation table are read together.
The field front-loaded its claims
A single peak year followed by a drop to 2 filings by the midpoint points to an early land-grab around core defect-prediction and process-control concepts, not a steadily maturing field. New entrants filing broad process-control claims now are more likely to run into prior art from the 2020-2021 wave than into an open field.
Defect prediction leans on vision pipelines
Image data processing appears in nearly half the dataset, while dedicated AI-model computing (G06N) appears in only 2 families. Claims tend to be written around applying vision-based analysis to a specific manufacturing step, which leaves the underlying model architecture itself comparatively open.
Coverage is concentrated, not global
Europe and the United States account for the large majority of receiving offices, with only 2 PCT filings and single filings in China and Germany. Applicants targeting other manufacturing hubs may find less crowded prior art there.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to electron beam melting ai and machine learning, with the prior art for and against each one.
Who is filing, and where the gate sits
Momentum among the named assignees shows zero activity in the latest tracked year across the group, consistent with the broader flat-to-declining trend after 2021 — this is a field where the earliest filers set the boundaries and later entrants are working around them rather than racing to file first.
No assignee is currently accelerating
Every named assignee in the recent-momentum data shows zero filings in the latest tracked year, including entities with a -100% YoY change from a prior filing. This matches the dataset's overall flat-to-declining trend rather than indicating any single company has withdrawn from the space.
A small, concentrated dataset
With 26 total families, this is a narrow but well-defined niche rather than a broad technology area. That makes individual claim scope more consequential — a handful of well-drafted patents can meaningfully shape the freedom to operate.
Europe leads on receiving-office count
The EPO receives more filings in this dataset than the US, with WIPO PCT, China and Germany trailing well behind. Entities building a defensive position should weight European prior art review accordingly.
| Assignee | Recent year | YoY |
|---|---|---|
| Raytheon Technologies | 0 | -100% |
| Siemens Energy Global | 0 | — |
| Technology Research Association for Future Additive Manufacturing (TRAFAM) | 0 | — |
| General Electric Company | 0 | — |
| Siemens AG | 0 | — |
| The Boeing Company | 0 | — |
| Beijing University of Technology | 0 | — |
| University of Liverpool | 0 | — |
Where to take this analysis
The dataset points to a few concrete next steps for anyone drafting or reviewing filings in this space.
Check freedom to operate against the 2021 filing wave
Since the peak year concentrated a large share of families, prior-art searches should weight 2020-2022 filings heavily before assuming a claim direction is open.
Explore prior art in EurekaMap claims against the G06T-heavy vision pipelines
With image data processing present in nearly half the dataset, compare a proposed claim's sensor and vision approach against existing defect-prediction filings before drafting.
Run a claim comparison in EurekaWatch jurisdictions outside Europe and the US
With only 2 PCT filings and single filings in China and Germany, monitor those offices for new entrants rather than assuming coverage gaps will persist.
Track new filings in EurekaCommon questions about this landscape
The dataset includes filings from large industrial players such as GE, Siemens, Siemens Energy, Boeing, RTX and a Japanese next-generation additive manufacturing research consortium, alongside university filers. No single assignee dominates the 26-family dataset, and recent-year momentum data shows zero filings in the latest tracked year across all of these named assignees, which points to a field where early positions were staked out around 2021 rather than one with an active current leader. Anyone assessing competitive position should look at cumulative family counts and claim scope rather than recent filing velocity alone.
Filing activity peaked in 2021 at 13 families and had fallen to 2 by the 2022 midpoint, which is a flat-to-declining pattern rather than sustained growth. Because publication lags filing by around 18 months, the most recent one to two years in the trend are understated and should not be read as a confirmed drop-off. Taken together, the evidence suggests an initial wave of claiming activity that has not yet been followed by a second wave, though it is too early to rule one out.
The IPC composition shows additive manufacturing (B33Y, 13 families) and powder metallurgy (B22F, 12) as the core classes, with image data processing (G06T, 11) close behind, indicating that most defect-prediction claims are built around vision or sensor-data analysis applied to the powder-bed process. Control and regulating systems (G05B) and general digital data processing (G06F) each appear in 5 families, while AI-model computing specifically (G06N) appears in only 2. This means the field is dominated by process-and-sensor framed claims rather than claims centred on novel model architectures.
The thin representation of G06N relative to G06T suggests that claims focused on the underlying machine learning model architecture, rather than its application to vision data from a manufacturing process, remain comparatively under-claimed. Sub-areas such as in-situ melt pool sensor fusion, cross-layer defect propagation prediction, and model-based scan path replanning show limited direct coverage in the current IPC and citation data. A prior-art search focused specifically on those branches, rather than on the broader defect-prediction category, is likely to surface fewer blocking references.
Europe, through the EPO, and the United States together account for 21 of the 26 receiving-office filings in this dataset, making them the primary jurisdictions to clear before filing. WIPO PCT filings are limited to 2, and China and Germany each show only a single filing, indicating that broad international coverage is still the exception rather than the norm in this niche. Filers targeting manufacturing hubs outside Europe and the US may encounter less crowded prior art, though the small sample size means this should be verified with a direct search rather than assumed.
Research Electron Beam Melting AI and Machine Learning in depth with Eureka
Go past this page: query the whole electron beam melting ai and machine learning corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.