Robot Joint Torque Sensing Patents: Who Leads, Where the Gaps Are 2026
- 71.1% of all filings sit with five assignees. 27 of the 38 records in scope trace to the top five filers, with a long tail of single-record entrants behind them.
- Filing pace has cooled from its 2019 peak. 2019 recorded 8 filings; the 2021-to-2024 window shows a documented -50% change, though 2025 onward is still filling in as publications lag filing.
- Force measurement claims overlap manipulator claims in over half the field. B25J and G01L each cover more than half of the 38 records, meaning most torque-sensing claims are being written into the manipulator mechanics itself, not filed as standalone sensor patents.
Filing growth compares 2021 (2 records) with 2024 (1) — 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 38 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This landscape tracks patent filings on robot joint torque sensing — the strain gauges, flexible transducers and calibration methods that let a robot arm measure and correct the torque at its own joints. The scope spans 2015 to the 2026-07-31 data cut-off and captures 38 published records matched on torque-sensor terminology together with claim language on crosstalk between axes, temperature drift, stiffness, sensitivity, strain gauge bonding, overload protection or calibration.
Most of the activity sits at the intersection of manipulator mechanics and force measurement, with a smaller but persistent thread in surgical robotics. Filing has been dominated by a handful of research-heavy filers rather than a broad industrial base, and the receiving-office mix points to China as the primary jurisdiction of record.
Filing trend and technology composition
Two views of the same 38 records: how filing activity has moved year over year, and which IPC subclasses the claims actually sit in.
Filing trend: a 2019 peak, then a cooling patch
Filings ran from 2 in 2017 up to a peak of 8 in 2019, then eased off. The documented change from 2021 (2 filings) to 2024 (1 filing) is -50% over that three-year span — 2024 is the most recent year that can be read as complete, since publication typically lags filing by around 18 months and 2025-2026 records are still arriving.
IPC composition: manipulators and force measurement dominate
B25J (manipulators & robots) covers 57.9% of the 38 records and G01L (force & pressure measurement) covers 52.6% — the two overlap heavily, since a joint torque sensor is claimed as much as a mechanical structure as a measurement device. A61B (diagnosis & surgery) reaches 18.4%, tracing the surgical-robotics thread, while G16H healthcare informatics (7.9%) and G06F data processing (5.3%) are thin enough to represent adjacent rather than core claim territory.
Shares are the percentage of the 38 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Robot Joint Torque Sensing with Eureka
This page is one run against one query. Ask Eureka your own question about robot joint torque sensing and every answer comes back with the patent numbers behind it.
Try EurekaMost-cited records and a recent filing
WO2025031575A1 — intrinsic friction and torque ripple compensation via an integrated joint torque sensor
A method for measuring torque with a sensor integrated into a robot arm joint: a calibration step captures torque, position, velocity and acceleration during no-load rotation to build training data, then a linear regression step derives a correction function from at least four basis functions, including at least one frictional torque basis function with a velocity dependency.Filed by Schaeffler Technologies AG & Co. KG, published 2025-02-13 — one of the most recent records in scope and a useful marker of where active claim drafting is happening now.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | CN104215372A | 一种机械臂关节扭矩测量装置 | 21 |
| 2 | US20220233271A1 | System and apparatus for external torque observation and compensation for surgical robotic arm | 20 |
| 3 | CN108709683A | 面向模块化关节的柔性力矩传感器动态测试平台 | 14 |
| 4 | CN216884037U | 一种机器人关节模组 | 13 |
| 5 | CN108709684A | 面向模块化关节力矩传感器的自主标定平台 | 10 |
| 6 | CN112665765A | 一种基于并联分载原理的机器人高刚度关节力矩传感器 | 9 |
| 7 | CN108931328A | 一种机器人关节扭矩传感器 | 7 |
| 8 | CN112525420A | 一种机器人关节力矩传感器自动标定装置 | 6 |
| 9 | WO2020246997A1 | System and apparatus for external torque observation and compensation for surgical robotic arm | 6 |
| 10 | CN114577381A | 一种机器人关节扭矩传感器及其扭矩测量方法 | 5 |
Citation counts inside a searched corpus favour older filings that have had more time to accumulate citers — read this as a signal of influence within the corpus, not a ranking of current technical relevance.
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 data means for filing strategy
Three read-throughs from the concentration, trend and classification figures above.
The field is led, not fragmented
Five assignees account for 27 of the 38 records in scope. That is a narrow leading group for a 16-assignee ranking, with the leader alone holding 12 records — new entrants are filing into claim space that a small number of players have already mapped.
Activity has thinned since the 2019 peak
The 2019 peak of 8 filings has not been repeated. The documented 2021-to-2024 change is a decline, though the 2025-2026 years in the dataset are still incomplete because publication lags filing by roughly 18 months, so any read of the very latest years understates true activity.
Claims are written into the mechanism, not just the sensor
More than half of records fall in B25J and more than half in G01L, and the overlap between them is substantial. Drafting a torque-sensing claim that avoids both classes is difficult; most viable claim space is in how the two are combined — bonding geometry, signal routing, calibration method — rather than in either domain alone.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to robot joint torque sensing, with the prior art for and against each one.
Who is filing, and where the collaboration sits
A short leading group, a long tail of single-record filers, and one notable co-filing pair.
One filer holds the largest single block
The leading assignee accounts for 12 of the 38 records in scope, well ahead of fifth place at 2 records — a gap that suggests one organisation has built a defensible position across multiple filings rather than a single landmark patent.
One recurring collaboration stands out
Schaeffler Technologies and Nanyang Technological University co-appear on 3 records, the strongest co-assignee pair in the dataset — evidence of a sustained joint research programme rather than a one-off filing.
Recent-year activity has gone quiet across the board
Among tracked assignees, the most active in the latest year shows only 1 filing at flat year-on-year change; several previously active filers, including the corpus leader, show zero filings in the latest year. This reads more as a publication-lag gap than a genuine pullback.
| Assignee | Recent year | YoY |
|---|---|---|
| Dalian University of Technology | 1 | 0% |
| Harbin Institute of Technology | 0 | — |
| Covidien LP | 0 | -100% |
| Schaeffler Technologies AG & Co. KG | 0 | — |
| Nanyang Technological University | 0 | — |
| Korea Institute of Science and Technology | 0 | — |
| Hangzhou Dianzi University | 0 | — |
| KUKA Deutschland GmbH | 0 | — |
Where to take this analysis
The dataset points to a concentrated leading group and several thin branches worth a closer look before committing a filing strategy.
Check freedom-to-operate against the leading assignee
With one filer holding 12 of 38 records, any new joint-torque-sensor filing should be checked claim-by-claim against that portfolio before drafting.
Run a claim comparison in EurekaMap the co-filing network further
The Schaeffler–Nanyang pairing suggests joint academic-industry programmes are active in this space; tracing similar pairs elsewhere may surface upcoming filings before they publish.
Explore assignee networks in EurekaWatch the under-claimed branches
Crosstalk compensation, temperature-drift correction and autonomous calibration platforms all show thinner coverage than the B25J/G01L core — each is a candidate for a first-mover filing.
Search white space in EurekaCommon questions on robot joint torque sensing patents
The dataset ranks 16 assignees, with a single leader holding 12 of the 38 records in scope — far ahead of fifth place at 2 records. The top five filers combined account for 71.1% of all records, so the field is led by a small group rather than evenly spread. Beyond the top ten, who together hold 92.1% of records, filing activity thins into a long tail of single-record entrants, mostly universities and research institutes.
Filing peaked at 8 records in 2019 and has not returned to that level since. The documented change from 2021 (2 filings) to 2024 (1 filing) is a -50% shift, and 2024 is the most recent year that can be treated as complete. Filings from 2025 onward are still understated in any dataset because publication typically lags the original filing date by around 18 months, so a slowdown reading should stop at 2024, not extend into the most recent years shown.
The two dominant IPC subclasses are B25J (manipulators and robots), covering 57.9% of the 38 records, and G01L (force and pressure measurement), covering 52.6%. Because a single record can carry both classes, claims in this space are typically drafted at the intersection — a mechanical joint structure with an integrated force-measurement function — rather than as a standalone sensor. A61B (diagnosis and surgery) reaches 18.4% of records, marking a distinct surgical-robotics sub-thread, with healthcare informatics and data-processing classes appearing more thinly.
Relative to the dense B25J/G01L overlap, sub-areas such as cross-axis crosstalk compensation, temperature-drift correction, overload protection circuitry, strain-gauge bonding geometry and autonomous self-calibration platforms show thinner direct coverage. These are drawn from the search terms defining this dataset's scope, and none appears as a dominant standalone IPC class, suggesting claims here are still often bundled into broader joint-mechanism patents rather than filed on their own. That gap is where a narrowly drafted first claim is more likely to clear prior art.
WO2025031575A1 was filed by Schaeffler Technologies AG & Co. KG and published on 2025-02-13. It describes a calibration method for a torque sensor integrated into a robot arm joint: a no-load calibration step captures torque, position, velocity and acceleration to build training data, then linear regression against at least four basis functions — including a velocity-dependent frictional torque function — derives a correction for the joint's intrinsic dynamics. It is one of the most recent filings in this dataset and marks where active claim drafting sits today, though its scope should be checked in full against a live patent record before relying on it for freedom-to-operate decisions.
Research Robot Joint Torque Sensing in depth with Eureka
Go past this page: query the whole robot joint torque sensing 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.