Soft-Sensor Bioprocessing Patents: Who Leads, Where the Gaps Are 2026
- 68.7% concentration. The top five assignees together hold 365 of 531 records in scope (68.7%), with one filer alone accounting for 274 — this field is not a level playing field for a new entrant.
- Growth has held, not stalled. Filings rose 46 to 48 between 2021 and 2024 (+4%), and 2025-26 figures look lower only because publication lags filing by roughly 18 months.
- Control and AI classes dominate together. G05B (control systems) and G06N (AI models) each cover roughly three-quarters of records, meaning most filings pair sensor logic with a learning method rather than claiming either alone.
Filing growth compares 2021 (46 records) with 2024 (48) — 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 531 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This landscape tracks patent activity at the intersection of soft-sensor and virtual-sensor methods and the variable-selection problem that makes them work in practice: choosing which process signals to trust, screening for redundant or lagged inputs, and handling data quality before a model ever sees the numbers. The search combines soft-sensor terminology with feature-selection and sensor-redundancy language, scoped to control-systems, bioprocess-assay and machine-learning IPC classes. It captures 531 published records filed between 2015 and the 2026 cut-off.
Most applicants come from industrial-IoT and process-control backgrounds rather than pure bioprocess houses, which shapes where the claim density sits: heavy on sensor-fusion and classifier architecture, lighter on bioprocess-specific assay chemistry. That split matters for anyone assessing freedom to operate in a bioprocessing-specific application.
Filing trend and technology composition
Two views of the same 531-record dataset: how filing activity has moved year over year, and which IPC subclasses carry the claims.
Filing trend, 2017-2026
Filings climbed to a peak of 120 in 2019, then settled into a steadier band; the 2021-to-2024 comparison (46 to 48, +4%) is the most reliable recent read since 2025-26 counts are still being filled in by publication lag.
Technology composition by IPC subclass
G05B (control and regulating systems, 78.9% of records) and G06N (AI-based computing, 76.8%) are the two largest classes and heavily overlap, since a record can carry both; H04L and H04B transmission classes each cover roughly half the dataset, reflecting how much of this art is really about getting sensor data somewhere reliably before selecting a variable.
Shares are the percentage of the 531 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Soft-Sensor Bioprocessing Variable Selection with Eureka
This page is one run against one query. Ask Eureka your own question about soft-sensor bioprocessing variable selection and every answer comes back with the patent numbers behind it.
Try EurekaA representative claim and the most-cited prior art
US11656115B2 — Method for determining a process variable with a classifier for selecting a measuring method
The method records a first value for a process variable using one measuring method and a second value using a second method, then uses a classifier to select which recorded value to output as the process variable. It also covers a computer program and system configured to execute the method.Filed by Endress+Hauser SE+Co. KG, granted 2023-05-23.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20190339688A1 | Methods and systems for data collection, learning, and streaming of machine signals for analytics and mainten… | 973 |
| 2 | US20200348662A1 | Platform for facilitating development of intelligence in an industrial internet of things system | 732 |
| 3 | US20210157312A1 | Intelligent vibration digital twin systems and methods for industrial environments | 715 |
| 4 | US6207936B1 | Model-based predictive control of thermal processing | 579 |
| 5 | US20200225655A1 | Methods, systems, kits and apparatuses for monitoring and managing industrial settings in an industrial inter… | 563 |
| 6 | US20200103894A1 | Methods and systems for data collection, learning, and streaming of machine signals for computerized maintena… | 513 |
| 7 | US20190171187A1 | Methods and systems for the industrial internet of things | 457 |
| 8 | US20090198350A1 | Robust adaptive model predictive controller with tuning to compensate for model mismatch | 291 |
| 9 | US20200327371A1 | Intelligent Edge Computing Platform with Machine Learning Capability | 286 |
| 10 | US20210248514A1 | Artificial intelligence selection and configuration | 279 |
Citation counts favour older records in any searched corpus; treat this table as a map of influential prior art, not of current filing activity.
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 read-outs from the dataset that matter more for strategy than the raw counts alone.
The field is led, not fragmented
One assignee holds 274 of 531 records outright, and the top five together hold 68.7% of all records in scope. A new entrant is filing into a space where the largest claim positions are already staked, particularly around industrial-IoT sensor architectures.
Steady, not accelerating or collapsing
Filings moved from 46 in 2021 to 48 in 2024, a modest but real increase. The apparent drop in 2025-26 numbers is a publication-lag artefact, not a sign the field has gone quiet.
Variable selection is now an ML problem
Three in four records carry an AI-computing classification alongside control-systems classes, confirming that variable and feature selection in this field is being claimed as a learned function rather than a fixed rule set.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to soft-sensor bioprocessing variable selection, with the prior art for and against each one.
Who is filing, and where momentum has moved
The ranked leaders come mostly from industrial-IoT and process-automation backgrounds, with process-instrumentation and pharma names holding smaller but focused positions.
Industrial-IoT platform filers dominate by volume
The leading assignee's filings sit largely in machine-signal collection and industrial-IoT analytics rather than bioprocess-specific sensing, meaning its claim breadth is wide but not necessarily bioprocess-native.
Process automation names hold the next tier
Names like Honeywell, Fisher-Rosemount and ABB round out the top ten, bringing decades of process-control patent estate into a field increasingly framed around ML-based selection.
Recent activity has cooled across the leaders
Several of the most active historical filers, including the top-ranked assignee, show zero filings in the latest tracked year; this likely reflects publication lag rather than an actual pullback.
| Assignee | Recent year | YoY |
|---|---|---|
| STRONG FORCE TX PORTFOLIO 2018 LLC | 2 | -33% |
| Strong Force IoT Portfolio 2016, LLC | 0 | -100% |
| ISTARI DIGITAL INC | 0 | -100% |
| Tata Consultancy Services Ltd. | 0 | — |
| Fisher-Rosemount Systems, Inc. | 0 | — |
| Honeywell International Inc. | 0 | — |
| F. Hoffmann-La Roche AG | 0 | — |
| ABB (Schweiz) AG | 0 | — |
Where to take this next
The dataset points to a field with a settled leader tier and open ground around bioprocess-specific applications of variable selection.
Map freedom to operate against the leader tier
Run a focused clearance search against the top five assignees before drafting claims in sensor-fusion or classifier-based variable selection.
Explore in EurekaTrack the under-claimed branches
Time-lag correction and cross-batch input screening show filing activity without a dominant holder — worth monitoring for a first-mover claim.
Explore in EurekaWatch the 2025-26 filings as they publish
Because of the 18-month publication lag, the most recent two years will keep filling in; revisit the trend once 2025 data settles.
Explore in EurekaCommon questions about this landscape
It is fairly concentrated: the top five assignees together hold 68.7% of the 531 records in scope, and the single largest filer accounts for 274 of those records alone. The top ten combined reach 78.0% of all records. That leaves a long tail of smaller filers holding the remaining share, so a newcomer's freedom to operate depends heavily on clearing the largest holders first.
Filings grew from 46 in 2021 to 48 in 2024, a modest +4% increase over that span, and the field peaked at 120 filings in 2019. Numbers for 2025 and 2026 appear lower, but that is expected: patent publication typically lags filing by around 18 months, so those years are still filling in rather than showing an actual slowdown.
Control and regulating systems (IPC class G05B) and AI-based computing (G06N) are the two largest classes, covering 78.9% and 76.8% of the 531 records respectively. Digital transmission classes (H04L, H04B) also cover roughly half the dataset, reflecting how much of the claimed art concerns getting sensor data reliably to a model before any variable is selected. Because records can carry multiple IPC classes, these shares overlap and add to more than 100%.
The ranked leaders are dominated by industrial-IoT and process-automation companies rather than bioprocess-specific players, with the largest single filer's estate built around machine-signal analytics for industrial settings. Established process-control and instrumentation names such as Honeywell, ABB and Fisher-Rosemount hold meaningful positions in the next tier. Pharma and process-instrumentation names like Roche and Endress+Hauser appear with more narrowly bioprocess-relevant claims.
The clearest openings sit in bioprocess-specific applications of variable selection that the industrial-IoT leaders have not claimed tightly: time-lag correction tuned to fermentation dynamics, sensor-redundancy scoring for single-use bioreactors, and cross-batch input screening for scale-up. These branches show filing activity in the dataset but no single dominant holder, unlike the heavily claimed general sensor-fusion and classifier architectures.
Research Soft-Sensor Bioprocessing Variable Selection in depth with Eureka
Go past this page: query the whole soft-sensor bioprocessing variable selection 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.