BIM AI & Machine Learning Patents: Leaders, Trends & Gaps 2026
- Filing activity has plateaued, not grown. The dataset peaked at 15 families in 2022, and later years sit below that midpoint — a flat trajectory rather than the accelerating curve you'd expect from an AI-adjacent field.
- Business-process claims rival core software claims. G06Q (business, commerce & admin data processing) appears in 36 of 55 families, nearly matching G06F's 43 — a large share of this art is about workflow and cost, not just model architecture.
- Co-filing is rare and concentrated among three names. Only five co-assignee pairs exist across the whole dataset, and the strongest pairings all involve the same three organizations — most applicants file alone.
Filing growth compares 2021 (3 records) with 2024 (10) — a three-year span. 2024 is the most recent year we treat as complete: publication lags filing by roughly 18 months, so 2025 onwards are still filling in and any growth rate that ends there would understate the field. Top-5 share is the combined record count of the five largest assignees divided by all 55 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This dataset tracks patent families at the intersection of building information modeling (BIM) and machine learning-driven construction workflows: generative design inside BIM platforms, automated clash resolution, construction-progress AI, and the underlying G06N/G06Q/G06F classification stack. It spans filings from 2015 through the 2026 cut-off, with the earliest matched record in 2017 and 55 total families.
Because publication lags filing by roughly eighteen months, the most recent one or two years in any trend line will always look thinner than they eventually turn out to be. Read the 2025-2026 tail as a floor, not a ceiling.
Filing trend and technology composition
Fifty-five families, six receiving offices, and a technology mix that leans heavily on general-purpose data processing classes rather than narrow AI-only codes.
A plateau, not a climb
Filings rose from a single family in 2017 to a peak of 15 in 2022, then eased off. With 2022 as the midpoint of the observed range, the second half of the window has not exceeded the first — this is a field that expanded quickly once and has not repeated the move.
Where the claims sit
G06F (electric digital data processing) and G06Q (business, commerce & admin data processing) dominate at 43 and 36 records respectively, with G06T (image data processing) at 21 and G06N (AI computing) at 16. The gap between G06F/G06Q and G06N suggests most filers are patenting the BIM workflow and its outputs, with the AI model itself often a supporting element rather than the headline claim.
Shares are the percentage of the 55 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Building Information Modeling AI and Machine Learning with Eureka
This page is one run against one query. Ask Eureka your own question about building information modeling ai and machine learning and every answer comes back with the patent numbers behind it.
Try EurekaThe most-cited records in this space
Machine learning system for building renderings and building information modeling data
Techniques are disclosed for using a computation engine executing a machine learning system to generate, according to constraints, renderings of a building or building information modeling (BIM) data for the building, wherein the constraints include at least one of an architectural style or a building constraint. An input device receives an input indicating one or more surfaces for the building and one or more constraints; a machine learning system, trained using images of buildings labeled with corresponding constraints, applies a model to the surfaces to generate the rendering or BIM data.Filed by Obayashi Corporation, published 2022-10-11 as US11468206B2.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20220391627A1 | Rapid and accurate modeling of a building construction structure including estimates, detailing, and take-off… | 62 |
| 2 | US20200250354A1 | Building information modeling system with self-configuration | 30 |
| 3 | EP4339857A1 | System and method for automatically calculating quantity and construction cost of railway facilities based on… | 16 |
| 4 | US20180374052A1 | Method of improviding building information modeling | 15 |
| 5 | WO2021068061A1 | System and method for generating 3D models from specification documents | 14 |
| 6 | JP2020030786A | Machine learning system for building renderings and building information modeling data | 14 |
| 7 | US20220215135A1 | Systems And Methods For Training Modelers In Journey-Working | 12 |
| 8 | US20220391553A1 | Building information modeling systems and methods | 11 |
| 9 | US20230020885A1 | Automatic conversion of 2d schematics to 3D models | 10 |
| 10 | US20220383231A1 | Generating status of construction site based on hierarchical modeling that standardizes physical relationship… | 9 |
Citation counts inside this corpus favour older filings that have had more time to accumulate references — read them as a signal of influence on the field, not of current technical relevance.
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 data implies for filing strategy
Three patterns matter more than the raw counts: where citations concentrate, how thin the co-filing network is, and which classification pairs are actually being claimed together.
Influence sits with a handful of early filings
The most-cited record, on AI-driven modeling with cost and quantity take-offs, draws far more citations than the rest of the set. Two other top records date to 2018-2020, meaning the field's most-referenced ideas are already several years old and any new filing in that exact space is working against established prior art.
Most applicants file solo
Across 55 families there are only five recorded co-assignee pairings, and the strongest ones repeat the same three organizations. That points to a field where partnerships between research institutes, platform vendors and construction firms are still the exception, not standard practice.
Filing is US-anchored with a scattered international tail
The United States accounts for the largest share of receiving offices, with China, Europe and WIPO each in the high single digits and India and Japan close behind. No single non-US office has pulled ahead, suggesting applicants are not yet converging on a preferred second jurisdiction.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to building information modeling ai and machine learning, with the prior art for and against each one.
Who is active, and where the field is still open
No assignee shows filing growth in the latest recorded year — every organization with tracked momentum data reports zero filings in that period, including one with a full year-on-year drop. That is consistent with a field in a filing pause rather than a race between named leaders.
Recent-year activity has stalled across the board
Every assignee tracked for recent momentum — spanning platform vendors, a research institute, and specialist construction-AI firms — filed zero new families in the latest year, with one showing a full year-on-year decline. This is unusual for an AI-adjacent category and suggests the current filing wave has already crested.
A small cluster of repeat collaborators
The strongest co-assignee pairing in the dataset links a research institute with a generative-design platform vendor, and the same platform vendor also pairs with a Japanese general contractor. These three names account for most of the collaboration signal that exists at all.
Influence is held by a mix of vendors and system patents
The two most-cited records — one on AI-driven modeling with cost estimation, one on a self-configuring BIM system — set the reference points that later filings in the space tend to cite against, regardless of which organization filed them.
| Assignee | Recent year | YoY |
|---|---|---|
| DALUX APS | 0 | — |
| SRI International | 0 | — |
| INERTIA SYST | 0 | — |
| Exis Corporation | 0 | — |
| HYPAR INC | 0 | — |
| DOXEL INC | 0 | -100% |
| Obayashi Corporation | 0 | — |
| ZENDA LLC | 0 | -100% |
Where to take this analysis
The dataset points to a field that expanded once, in 2022, and has not repeated the move — the open questions now are about where claim density is lowest and which citation anchors still hold.
Map the white space against your own portfolio
Run a freedom-to-operate check against the under-claimed branches above, particularly where B25J, G06V and G09B overlap with core BIM claims — these classes carry a fraction of the filings that G06F and G06Q do.
Explore white space in EurekaTrack the assignees with stalled momentum
Every tracked assignee shows zero filings in the latest year. Watch for whether that reverses once the publication lag clears, or whether it signals a genuine pull-back from patenting in this space.
Monitor assignee activity in EurekaCommon questions about BIM AI and ML patents
This dataset identifies 55 patent families matching building information modeling combined with AI or machine learning claims, filed between 2015 and the 2026 cut-off with the earliest matched record from 2017. Filing peaked at 15 families in 2022 and has not exceeded that level since, so growth in this specific claim area is flat rather than accelerating. Because publication lags filing by around eighteen months, the 2025-2026 counts will rise somewhat as later filings publish, but the overall plateau pattern is unlikely to change.
The most-cited individual filings come from a mix of AI-modeling vendors and established contractors, including a highly-cited record on AI-driven construction estimating and another on a self-configuring BIM system from an established filer. Collaboration is limited: only five co-assignee pairs exist across the whole dataset, with a research institute, a generative-design platform vendor, and a Japanese general contractor forming the strongest links. No single assignee shows filing growth in the most recent tracked year, which points to a broad pause across active filers rather than one clear leader pulling ahead.
Classification data shows G06F (electric digital data processing) in 43 of 55 records and G06Q (business, commerce and admin data processing) in 36, well ahead of G06N (AI computing) at 16. That gap indicates most filers are claiming the BIM workflow, cost estimation, or data-processing pipeline around the AI, rather than the underlying machine learning model itself as the primary invention. Image processing (G06T, 21 records) and image/video recognition (G06V, 7 records) form a secondary cluster tied to progress monitoring and clash detection use cases.
The classes with the fewest matched records — B25J (manipulators and robots) at 1, G09B (educational and demonstration aids) at 2, and G06K (data recognition and presentation) at 3 — sit far below the G06F and G06Q clusters and represent comparatively open claim space. Concretely, that includes robotic execution of clash-resolution decisions, BIM-linked training simulators, and recognition systems bridging as-built imagery with as-designed models. A first claim in these areas would need to tie the under-claimed function directly to BIM data structures to avoid falling back into the crowded G06F/G06Q territory.
Not on this evidence. Filing rose from a single family in 2017 to a peak of 15 in 2022, and the years since have not surpassed that midpoint, making 2022 the high point of the observed window so far. Recent-year momentum data reinforces this: every assignee tracked for the latest year reports zero new filings, including one with a full year-on-year decline. The honest read is a field that had one strong filing wave and has since gone quiet, pending whatever the publication lag eventually reveals for 2025-2026.
Research Building Information Modeling AI and Machine Learning in depth with Eureka
Go past this page: query the whole building information modeling 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.