LPBF AI/ML Process Patent Landscape 2026
The AI/ML-driven laser powder bed fusion (LPBF) process space is in a confirmed growth stage, with annual filings expanding 78% over the recent window and the corpus reaching 190 patent families. The field is fragmented at the top — the five largest filers account for only 16% of the hundred largest filers’ combined total — with China-based institutions and a US national laboratory trading the leading positions.
Lawrence Livermore leads a fragmented, fast-growing field
Lawrence Livermore National Security LLC holds the top position with 10 patent families, followed by Optalysys at 6 and Nikon SLM Solutions AG at 5. The next tier — AECC Commercial Aircraft Engine, Nanjing University of Aeronautics and Astronautics, Huazhong University of Science and Technology, Tianjin University, and Taiwan Semiconductor Manufacturing — each hold 4 patent families, indicating a broadly distributed competitive base.
The top five filers together account for 16% of the hundred largest filers’ combined total, a notably low concentration figure. No single player has established dominant coverage, and the gap between rank 1 and rank 5 is narrow (10 vs. 4 patent families), signaling that the competitive hierarchy remains highly contestable.
| # | Applicant | Patent families | Share |
|---|---|---|---|
| 1 | Lawrence Livermore National Security LLC | 10 | |
| 2 | Optalysys | 6 | |
| 3 | Nikon SLM Solutions AG | 5 | |
| 4 | Bull SA | 4 | |
| 5 | AECC Commercial Aircraft Engine Co. Ltd. | 4 | |
| 6 | NANJING UNIV OF AERONAUTICS & ASTRONAUTICS | 4 | |
| 7 | AECC Shanghai Commercial Aircraft Engine Manufacturing Co. Ltd. | 4 | |
| 8 | HUAZHONG UNIV OF SCI & TECH | 4 | |
| 9 | Tianjin University | 4 | |
| 10 | Taiwan Semiconductor Manufacturing Co. Ltd. | 4 |
| # | Applicant | Patent families | Share |
|---|---|---|---|
| 11 | Wuhan University | 3 | |
| 12 | NANJING UNIV OF AERONAUTICS & ASTRONAUTICS WUXI RE… | 3 | |
| 13 | Xi’an Jiaotong University | 3 | |
| 14 | PredictSpring | 3 | |
| 15 | South China University of Technology | 3 | |
| 16 | RTX Corporation | 3 | |
| 17 | CHONGQING INST OF GREEN & INTELLIGENT TECH CHINESE… | 3 | |
| 18 | Xidian University | 3 | |
| 19 | Google LLC | 3 | |
| 20 | Siemens AG | 3 |
Lawrence Livermore’s lead reflects a government-funded R&D mandate in process qualification and materials characterization, while the cluster of Chinese universities and aerospace engine manufacturers points to industrial adoption priorities in aviation-grade components. New entrants can still stake meaningful positions without competing against entrenched patent thickets.
Patent filings from approximately 2024–2026 are under-represented due to typical publication lags of 12–18 months; the most recent annual figures should be treated as floor estimates rather than final counts. Longer-window growth, applicant concentration, and technology-route coverage are therefore more reliable signals than the latest-year bar alone.
Rapid growth since 2021, anchored by additive manufacturing and AI computing classes
Annual filing volume and technology class distribution together reveal a field still in active expansion, with the AI and additive manufacturing branches developing in parallel rather than sequentially.
Annual filing trend
Filings grew steadily from 2 in 2017 to a recorded high of 47 in 2025, with a 78% increase over the recent window. The 2026 figure of 4 reflects publication lag and is not a signal of slowdown. A visible dip in 2022 (9 filings) interrupted the growth trajectory before a strong rebound in 2023 (28) and continued acceleration through 2024–2025.
↗ Hover for values · click a bar to ask EurekaTechnology composition
Three branches — B33Y (additive manufacturing/3D printing, 107 records), B22F (powder metallurgy, 106 records), and G06N (AI computing models, 105 records) — are nearly co-equal in coverage, confirming that this corpus genuinely bridges process engineering and machine learning. G06F (digital data processing, 46 records) and G06V (image/video recognition, 24 records) form a secondary tier relevant to in-situ monitoring and quality inspection applications. Lower-share branches such as G05B (control and regulating systems, 9 records) and G01N (material analysis and testing, 8 records) remain sparsely covered.
↗ Hover for values · click a bar to ask EurekaHighly cited patent families surfaced by the query
Citation-heavy patent families returned by the query. Use this section as citation context, not as a curated list of the most topic-specific patents.
Machine learning thermal management system for add…
A thermal measurement system enables high-resolution sub-surface temperature monitoring during additive manufacturing processes through machine learning demodulation of chirped fiber Bragg grating (C-FBG) sensors. An optical sensing subsystem includes a C-FBG sensor that encodes spatial temperature information in wavelength for high-temperature operation. A… (excerpt from the patent abstract)
| # | Patent | Citations |
|---|---|---|
| 1 | Neural processor with holographic optical paths an… | 88 |
| 2 | 一种针对激光选区熔化成形缺陷的在线检测与优化系统 | 50 |
| 3 | Optical processor for an artificial neural network | 42 |
| 4 | 基于神经网络的ACO-OFDM系统综合PAPR抑制方法及系统 | 35 |
| 5 | Method for automatically preventing defects potent… | 25 |
| 6 | Neural processor with holographic optical paths an… | 24 |
| 7 | Predicting system in additive manufacturing proces… | 23 |
| 8 | 一种激光选区熔化技术原位质量综合评价方法 | 19 |
Ranked by total forward citations. Citation counts favour older and broadly cited patent families, and broad or adjacent patents may appear when they match the search scope. Treat this section as citation context, not as a curated list of the most topic-specific patents. Some patent titles may be shown in their original, non-English language where an accurate translation could not be guaranteed.
What the patent structure means for R&D investment decisions
The combination of low concentration, broad geographic filing, and strong annual growth defines a field where strategic positioning is still achievable. The following cards draw out the key structural signals.
Growth stage: annual filings still rising
The lifecycle evidence classifies this field as Growth, with annual filings still rising and recent-window growth of 78%. The 2024 and 2025 cohorts are the largest on record, and the 2026 figure is artificially suppressed by publication lag. Teams entering now will face a moving frontier rather than an entrenched prior-art wall, but the window for low-cost freedom-to-operate is narrowing as institutional filers accelerate.
Growth stageLow concentration: top five hold 16% of the leading hundred filers
The top five applicants collectively account for 16% of the hundred largest filers’ combined total — an unusually low figure for a maturing technical field. The gap between the leader (10 patent families) and the fifth-ranked applicants (4 patent families each) is small enough that a focused filing program over 12–18 months could move a new entrant into the top tier. No participant has yet built a blocking portfolio across the full AI-LPBF stack.
FragmentedCo-filing concentrated in Chinese aerospace and university pairs
The most active co-filing relationship (4 joint families) is between AECC Commercial Aircraft Engine Co. and AECC Shanghai Commercial Aircraft Engine Manufacturing Co., reflecting coordinated IP strategy within the same state enterprise group. Nanjing University of Aeronautics and Astronautics collaborates with its own Wuxi Research Institute (3 joint families), indicating intra-institution portfolio structuring. Huazhong University of Science and Technology and Xi’an Aerospace Engine Co. share 1 joint family. Cross-sector or international co-filing is not yet in evidence at scale.
Ecosystem formingChina dominates filings; US and PCT routes provide international coverage
China accounts for the largest share of patent records, followed by the United States (43 records) and WIPO PCT (22 records). Europe via the EPO (14 records) and India (11 records) form a secondary tier. The UK (5 records), South Korea (3 records), Germany (2 records), and Taiwan (2 records) are present but sparse. Japan holds only 1 record despite being a major LPBF equipment market — a gap that may reflect either licensing strategy or under-filing relative to commercial activity.
China-ledGo beyond the landscape: Eureka’s TRIZ Solution agent breaks down an R&D problem and returns patented concept solutions, each with a technical approach and cited patent & literature evidence.
| Applicant | Collaborator | Co-filings |
|---|---|---|
| AECC Commercial Aircraft Engine Co. Ltd. | AECC Shanghai Commercial Aircraft Engine Manufacturing Co. Ltd. | 4 |
| Nanjing University of Aeronautics and Astronautics | Nanjing University of Aeronautics and Astronautics Wuxi Research Institute | 3 |
| Huazhong University of Science and Technology | Xi’an Aerospace Engine Co. Ltd. | 1 |
Co-filing pairs, ranked by the number of jointly-filed patent families.
Lawrence Livermore and Optalysys lead by distinct technology routes
The top two applicants reflect fundamentally different approaches: one anchored in physical process control for LPBF, the other in optical neural network hardware that intersects with the field. Trajectory data marks both as new entrants on a multi-year basis.
Lawrence Livermore National Security LLC
Holds 10 patent families, the largest position in the corpus. Technology emphasis is squarely on LPBF process engineering: the top focus areas are B33Y10 (additive manufacturing process, 9 records), B22F10 (powder metallurgy process, 7 records), and B33Y30 (additive manufacturing apparatus, 7 records). Momentum is flagged as a new entrant on a multi-year basis, meaning its current lead has been built recently rather than over a long filing history — suggesting the portfolio is concentrated in a short, intensive campaign rather than compounded over years.
patent families: 10Optalysys
Holds 6 patent families, ranked second. Its technology emphasis diverges sharply from the process-control mainstream: top focus areas are G06N3 (neural network computing, 6 records), G06V10 (image/video recognition, 6 records), and G06K9 (data recognition, 3 records) — indicating a positioning around optical computing hardware for AI inference rather than LPBF process control per se. This makes Optalysys a potential enabler or supplier to LPBF AI systems rather than a direct process IP competitor. Momentum is also flagged as a new entrant.
patent families: 6| Applicant | Recent (3 yrs) | Trend |
|---|---|---|
| Lawrence Livermore National Security LLC | 2 | ▲ new entrant |
| Hunan Luojia Intelligent Technology Co. Ltd. | 4 | ▲ new entrant |
| Tianjin University | 4 | ▲ new entrant |
| Taiwan Semiconductor Manufacturing Co. Ltd. | 3 | ▲ new entrant |
| Nanjing University of Aeronautics and Astronautics | 4 | ▲ new entrant |
| Huazhong University of Science and Technology | 1 | ▲ new entrant |
Under-served branches adjacent to the LPBF AI/ML core
Several IPC classes appear at the margins of the dominant B33Y/B22F/G06N core. These represent areas of relative sparsity and, in two cases, plausible technical value and realistic entry paths.
G05B · Control and regulating systems
With only 9 patent records, closed-loop real-time control systems — linking AI inference output directly to laser power, scan speed, or layer parameters — are sparsely covered relative to the field’s size. The technical case is strong: LPBF process quality is highly sensitive to moment-to-moment parameter adjustment, and AI models are increasingly capable of providing control signals in near-real-time. Existing work in G06N and G06V provides a platform from which a team with controls expertise could extend into G05B filings with limited prior-art overlap.
Search this in Eureka →G01N · Material analysis and testing
Only 8 patent records touch material analysis and testing (G01N), despite in-situ melt-pool characterization and post-build microstructure analysis being active research areas for AI-assisted LPBF qualification. The sparse coverage may reflect that most current work frames monitoring as image processing (G06V, G06T) rather than material science. A filing strategy that explicitly claims AI-driven methods for material property prediction or non-destructive evaluation of LPBF parts could occupy a relatively open adjacent space, particularly for aerospace and medical implant qualification workflows.
Search this in Eureka →How leading applicants differ across technology routes
Route coverage across the main technology branches in the current evidence set.
| Player | G06N 3 · Computing based on AI models | B22F 10 · Powder metallurgy | B33Y 50 · Additive manufacturing (3D printing) | B33Y 10 · Additive manufacturing (3D printing) | B22F 12 · Powder metallurgy |
|---|---|---|---|---|---|
| Lawrence Livermore National Security LLC | Absent | Strong · 7 | Strong · 7 | Strong · 9 | Strong · 6 |
| Nanjing University of Aeronautics and Astronautics | Strong · 3 | Strong · 3 | Strong · 3 | Strong · 2 | Absent |
| Tianjin University | Absent | Strong · 4 | Strong · 4 | Strong · 3 | Absent |
| Hunan Luojia Intelligent Technology Co. Ltd. | Absent | Strong · 4 | Strong · 4 | Absent | Moderate · 2 |
| Nikon SLM Solutions AG | Absent | Strong · 5 | Strong · 4 | Absent | Absent |
| Optalysys | Strong · 6 | Absent | Absent | Absent | Absent |
| AECC Shanghai Commercial Aircraft Engine Manufacturing Co. Ltd. | Absent | Absent | Strong · 3 | Strong · 3 | Absent |
Frequently asked questions
The corpus in scope contains 190 patent families. This figure covers global filings and is the basis for all share and ranking calculations in this report.
Lawrence Livermore National Security LLC holds the top position with 10 patent families, focused primarily on additive manufacturing process (B33Y) and powder metallurgy (B22F) classes. Optalysys ranks second with 6 patent families, emphasizing optical neural network computing.
The field is classified as Growth stage. Annual filings have risen with a 78% increase over the recent window, reaching recorded highs in 2024 and 2025. The 2026 figure is low due to publication lag and does not indicate a slowdown.
China is the leading filing jurisdiction, followed by the United States and WIPO PCT routes. Europe via the EPO and India form a secondary tier. Japan, despite being a major LPBF equipment market, holds only 1 patent record in the corpus — a notable gap.
Concentration is low. The top five filers account for 16% of the hundred largest filers’ combined total. The gap between the leader (10 patent families) and the fifth-ranked applicants (4 patent families each) is narrow, leaving the competitive hierarchy highly contestable.
The most-cited work includes a patent on neural processor with holographic optical paths (88 citations), a Chinese patent on online defect detection and optimization for selective laser melting (50 citations), and a patent on optical processor for artificial neural networks (42 citations). A patent on predicting systems in additive manufacturing processes also appears prominently (23 citations).
Ready to map your own LPBF AI/ML patent landscape?
Join 18,000+ innovators using PatSnap Eureka to map any technology landscape: search 2B+ patents and papers, surface key assignees, and generate a report like this in minutes.
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.