VRF System AI Patents: Who Leads, Where the Gaps Are 2026
- Filing activity peaked in 2019 at 8 records and has not returned to that level since, with the trend flat-to-declining through the midpoint year.
- F24F air-conditioning claims dominate at 20 of the corpus while G05B control-system claims (11) show these are largely control-layer inventions dressed in HVAC hardware language.
- The most-cited record, US20180004173A1, sits at 54 citations built around multi-level model predictive control — a strong signal of where the earliest defensible claim territory was staked.
Filing growth compares 2021 (3 records) with 2024 (0) — 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.
What this landscape covers
This dataset tracks patent families at the intersection of variable refrigerant flow (VRF) systems and machine learning fault detection, predictive control, energy optimization, and data-driven control, filtered to IPC classes covering air conditioning, machine learning, and refrigeration engineering. It spans filings from 2015 through the 2026 data cut-off, with 22 patent families forming the core of the ranking.
Publication typically lags filing by around 18 months, so the most recent filing years in any trend line understate real activity. Read the tail of the trend as a floor, not a ceiling.
Filing trend and technology composition
Two views of the same 22-family corpus: how filing activity has moved year over year, and which IPC subclasses carry the claim weight.
A peak in 2019, then a plateau
Filings rose from 3 in 2017 to a peak of 8 in 2019, then fell back; the midpoint year of 2022 recorded zero filings before activity picked up again. That pattern points to an initial wave of foundational control-system filings followed by a lull, rather than sustained compounding growth.
Control claims ride inside HVAC hardware classes
F24F (air conditioning & ventilation) covers 20 of the corpus, but G05B (control & regulating systems) appears in 11 — meaning most inventions here are control algorithms claimed through an HVAC-system wrapper rather than pure software filings. F25B, G05D, G06Q, F25D and H02J appear in smaller numbers, marking adjacent but thinner claim territory.
Shares are the percentage of the 22 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Variable Refrigerant Flow System AI and Machine Learning with Eureka
This page is one run against one query. Ask Eureka your own question about variable refrigerant flow system ai and machine learning and every answer comes back with the patent numbers behind it.
Try EurekaThe most-cited records
Variable refrigerant flow system with multi-level model predictive control
A model predictive control system optimizes energy cost in a VRF system composed of an outdoor subsystem and multiple indoor subsystems. A high-level model predictive controller generates optimal load profiles for each indoor subsystem to minimize energy cost, while low-level indoor MPCs generate optimal setpoints for the VRF units beneath them.Filed by Tyco Fire & Security GmbH, published 2018-01-04 — the most-cited record in this corpus at 54 citations.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20180004173A1 | Variable refrigerant flow system with multi-level model predictive control | 54 |
| 2 | US20190338973A1 | Variable refrigerant flow, room air conditioner, and packaged air conditioner control systems with cost targe… | 33 |
| 3 | US20230116964A1 | Systems and methods for controlling variable refrigerant flow systems and equipment using artificial intellig… | 24 |
| 4 | US11002457B2 | Variable refrigerant flow, room air conditioner, and packaged air conditioner control systems with cost targe… | 23 |
| 5 | US20230194137A1 | Systems and methods for controlling variable refrigerant flow systems using artificial intelligence | 12 |
| 6 | US10564612B2 | Variable refrigerant flow system with multi-level model predictive control | 10 |
| 7 | WO2021179250A1 | Systems and methods for controlling variable refrigerant flow systems and equipment using artificial intellig… | 8 |
| 8 | US20200393158A1 | Variable refrigerant flow system with zone grouping | 6 |
| 9 | US20210018204A1 | Variable refrigerant flow system with zone grouping control feasibility estimation | 4 |
| 10 | CN109416191A | 具有预测控制的变制冷剂流量系统 | 4 |
Citation counts are drawn from a searched corpus and skew toward older filings; treat them as a measure of influence on later filers, not of current commercial 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 say about this field
Three signals worth weighing before deciding where to file or who to watch.
A single wave, not a curve
The trend does not show steady compounding growth. It shows one concentrated wave of filings around 2019 followed by a gap. That is consistent with a handful of large HVAC controls players staking out core claims early, then pausing rather than a broad market ramping up continuously.
Control logic wrapped in HVAC hardware
Nearly all records touch F24F air-conditioning classification, but half also carry a G05B control-systems tag. Drafting here typically frames the invention as a VRF system claim with a control-loop dependent claim set, rather than a standalone algorithm claim.
US-centred prosecution
Seventeen of the tracked records were filed at the US receiving office, with four routed through the PCT system and one in China. Anyone assessing freedom to operate in this space should treat US prosecution history as the primary record to search.
Influence concentrated in early MPC filings
The top-cited record leads by a wide margin over the rest of the table, and the next two highest-cited records share the same underlying cost-target optimization concept across related filings. Later entrants are largely building on, or designing around, this small set of foundational claims.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to variable refrigerant flow system ai and machine learning, with the prior art for and against each one.
Who is filing, and where the field is still open
Assignee activity in this corpus is thin relative to the broader VRF market, with named individual inventors appearing alongside corporate assignees — a sign of a still-consolidating niche rather than a mature, heavily contested one.
No assignee shows recent-year momentum
Every assignee tracked in the recent-year momentum view, corporate and individual alike, recorded zero filings in the latest year. Given the roughly 18-month publication lag, this understates true activity but still suggests no single player is currently pulling ahead.
Co-filing is limited and inventor-led
The strongest co-assignee pair links two named inventors rather than two corporations, and the corporate co-filing pairs each appear only once. Collaboration in this space looks more like internal inventor teams than cross-company joint filing.
Tyco Fire & Security holds the anchor claim
The most-cited record in the corpus, on multi-level model predictive control for VRF energy cost optimization, was filed by Tyco Fire & Security GmbH. Its citation count sets it apart from the rest of the table and makes it the natural starting point for any freedom-to-operate review.
| Assignee | Recent year | YoY |
|---|---|---|
| Johnson Controls Technology Company | 0 | — |
| Johnson Controls Tyco IP Holdings LLP | 0 | — |
| WU ZHIGANG | 0 | — |
| FAN BO | 0 | — |
| WANG LUMING | 0 | — |
| SUN ZHONGYUE | 0 | — |
Where to take this from here
This landscape is a starting point for a freedom-to-operate check or a competitive scan, not a substitute for one.
Run a full-text claims comparison
Citation and IPC counts show where activity clusters, but only a claims-level read of the top-cited records will show exactly what is blocked versus what is open.
Open in Patsnap EurekaTrack the individual inventors
Several of the most active names in this corpus file as individuals rather than through a single corporate assignee, which makes ongoing monitoring by inventor name as useful as monitoring by company.
Set up inventor trackingCommon questions about this landscape
This landscape tracks 22 patent families published between 2015 and the mid-2026 data cut-off, filtered to VRF systems combined with machine learning fault detection, predictive control, energy optimization or data-driven control terms. That is a narrow, specific slice of the broader VRF patent space, not a count of every VRF-related filing. Because publication lags filing by roughly 18 months, the true number of filed-but-unpublished applications in the most recent year is higher than shown here.
The single most-cited record in this corpus, US20180004173A1, covers multi-level model predictive control for VRF energy cost optimization and was filed by Tyco Fire & Security GmbH; it carries 54 citations, well ahead of the rest of the table. Several of the other highly cited records share the same underlying cost-target optimization concept, filed under related VRF and room air conditioner control system titles. Citation counts favour older filings, so treat this as a map of foundational influence rather than current market share.
The trend is flat to declining rather than growing. Filings rose to a peak of 8 in 2019, then fell, with the midpoint year of 2022 recording zero tracked filings. That pattern is more consistent with an initial wave of foundational filings than with a technology in an active growth phase, though the recency of the corpus means the last one to two years are undercounted due to publication lag.
IPC composition shows heavy concentration in F24F (air conditioning and ventilation) and G05B (control and regulating systems), with much thinner coverage in F25B, G05D, G06Q, F25D and H02J. That distribution suggests areas such as refrigerant charge fault diagnosis, compressor degradation prediction, cross-unit load-balancing and demand-response cost-target control carry lighter claim density and are worth a fresh clearance search before filing. Thin coverage is a signal to investigate, not a guarantee the space is genuinely open.
Seventeen of the 22 tracked records were filed at the United States receiving office, with four filed through the PCT system and one in China. For most freedom-to-operate work in this field, US prosecution history should be the primary search, supplemented by the PCT family members to check for regional divergence in claim scope. A single China filing in this corpus is not enough to draw conclusions about that market's patent density in this niche.
Research Variable Refrigerant Flow System AI and Machine Learning in depth with Eureka
Go past this page: query the whole variable refrigerant flow system 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.