Preliminary Patentability Opinion – Warehouse Robots: Joint Torque Self‑Calibration Using Dual IMUs and Current Sensing, with Anomaly Circuit‑Breaker and Redundant Speed‑Limiting for Reliability and Maintenance Efficiency
I. Introduction and Framing of the Claimed Subject Matter
The application is understood to concern a control and diagnostics system for articulated warehouse robots in which:
- joint torque is estimated and self‑calibrated online using both motor current sensing and a pair of inertial measurement units (IMUs) positioned to sense kinematics across the joint (dual‑IMU configuration); and
- a safety and reliability layer uses anomaly detection to trigger a “fuse” (software/hardware circuit breaker) and redundant speed‑limiting mechanisms to place the robot into a safe degraded mode, thereby improving reliability and maintenance efficiency.
For purposes of this opinion, the independent claims are presumed to recite: (i) a method and/or apparatus implementing an online self‑calibration algorithm for joint torque constant and/or load/friction parameters using dual‑IMU measurements fused with motor phase current; and (ii) an integrated fault‑response framework that, on detecting discrepancies among sensor‑derived torque estimates, enforces a circuit‑breaker action and a speed‑limiting fallback through an independent channel.
II. Patent‑Eligible Subject Matter
A. United States (35 U.S.C. § 101)
- The claims are directed to the control of a physical robot joint using specific sensor configurations and control actions. Under Alice/Mayo, while mathematical relationships per se are abstract, claims that integrate such relationships into improved operation of a physical machine constitute a practical application.
- Controlling and calibrating a robot joint torque using an unconventional sensor arrangement has been held patent‑eligible when it improves accuracy and robustness of inertial tracking and control of real‑world systems (see, e.g., Diamond v. Diehr; Thales Visionix Inc. v. United States). The present subject matter, when claimed as integrated control of actuators and safety mechanisms, is likely eligible at Step 2A, Prong Two, or alternatively at Step 2B as reciting significantly more than an abstract idea.
- Recommendation: Emphasize concrete hardware constraints (sensor placement, sampling/synchronization, actuation interfaces), real‑time operation, and specific safety interlocks to preclude characterization as a mere mathematical algorithm.
B. Europe (EPC Art. 52(1))
- The invention is a computer‑implemented control of an industrial robot producing a measurable technical effect (improved torque calibration accuracy and fault response). Under established EPO practice (e.g., T 258/03, T 1227/05), such control solutions are technical. Therefore, subject‑matter eligibility is not expected to be at issue.
III. Novelty (35 U.S.C. § 102; EPC Art. 54)
General state of the art (without citing specific documents) includes:
- Joint torque estimation via motor current sensing and known torque constants; calibration of torque constants during commissioning; friction/gravity compensation; and observer/estimator‑based approaches (e.g., EKF/LS) using encoders and, in some instances, IMUs.
- Dual‑IMU configurations for measuring relative motion of articulated bodies and mitigating drift/bias in single‑IMU systems are known in robotics and motion tracking.
- Safety mechanisms such as speed limiting, watchdogs, fault detection using thresholding on residuals, and software “circuit breaker” patterns are also widely taught; functional safety frameworks for industrial and driverless/warehouse robots (e.g., ISO 13849‑1/‑2, IEC 61508, ISO 3691‑4, ISO 10218) contemplate redundant channels and degraded modes.
Given this baseline, the following features are the most likely to support novelty if explicitly recited and properly limited:
- A specific dual‑IMU placement across each joint with a defined fusion model that uses relative inertial data (e.g., joint‑local angular acceleration/jerk) to excite/observe torque‑related parameters and to self‑calibrate the motor torque constant and friction parameters in situ, during normal warehouse duty cycles, without external fixtures.
- A particular residual architecture that cross‑checks IMU‑derived joint dynamics against current‑based electromagnetic torque, with synchronization, bias estimation, and excitation/observability gating, and that selectively updates calibration only when predefined observability conditions are satisfied.
- An integrated anomaly response in which the same cross‑sensor residuals drive both (i) a circuit‑breaker action that decouples the calibrator from the control loop or disengages commanded torque output within a bounded reaction time, and (ii) an independent, redundant speed‑limiting path that remains effective even if the main controller is compromised.
Absent such specific interactions and constraints, the individual building blocks appear known. Accordingly, novelty hinges on the defined interplay, timing, gating, and architecture particulars.
IV. Inventive Step/Non‑Obviousness (35 U.S.C. § 103; EPC Art. 56)
A. Problem–Solution Approach (EPO)
- Closest prior art: A robotic joint controller that estimates torque using motor current and standard parameters; optionally an estimator using an IMU or encoder; and a safety system enforcing speed limits on fault.
- Objective technical problem: To reduce calibration drift and maintenance overhead in warehouse robots while increasing fault tolerance and operational safety, particularly under changing payloads and wear.
- Distinguishing features: If the claims specify (1) a dual‑IMU architecture across the joint with a defined fusion and synchronization scheme; (2) an online calibration routine that uses relative inertial measurements and motor current with explicit observability checks to update torque constant and friction models during normal operation; and (3) a diagnostics layer that uses the same residuals to actuate both a circuit‑breaker and an independent speed‑limiting path with specified diagnostic coverage and reaction times.
Assessment:
- Combining a current‑based estimator with IMU data to improve torque estimation would, at a high level, be an obvious aggregation if applied in a straightforward manner (KSR). However, non‑obviousness may be established if the applicant can show a synergistic integration that solves concrete implementation hurdles—e.g., dual‑IMU co‑calibration for bias drift, precise time‑alignment under variable duty cycles, model‑based observability gating to prevent mis‑calibration under low excitation, and a safety architecture that derives its triggers from multi‑sensor residuals and achieves a particular diagnostic coverage or performance level (e.g., PL d under ISO 13849‑1) without additional hardware.
- Redundant speed‑limiting and “circuit‑breaker” concepts are generally known. To be non‑obvious, the application should show how coupling these responses to the calibrator residuals—using multi‑threshold logic, hysteresis windows, and independence from the main control channel—achieves a measurable safety advantage with reduced nuisance trips and improved uptime.
B. U.S. Obviousness (KSR v. Teleflex)
- A mere juxtaposition of known elements (current sensors, IMUs, speed limits, circuit breakers) with predictable results is likely obvious. The claims may overcome § 103 where they recite:
- a particular dual‑IMU geometry and fusion algorithm that reduces bias/drift and improves torque constant identification beyond encoder‑only or single‑IMU systems;
- a defined excitation/observability regime and estimator (e.g., EKF with stated state vector including Kt, viscous/Coulomb friction, gravity parameters) that updates parameters only under validated conditions derived from residual statistics; and
- a safety interlock path physically/electrically or logically independent of the primary control processor, triggered by the residuals with specified thresholds, majority/consensus logic, and bounded reaction times that meet known safety standards.
- If evidence (e.g., test data) shows surprising robustness across payload variation and wear, with maintenance interval extension or quantifiable reduction of false trips, this would support non‑obviousness.
V. Clarity and Definiteness (35 U.S.C. § 112(b); EPC Art. 84)
Potential clarity issues:
- “Anomaly fuse” should be defined as a circuit‑breaker function, with indication whether it is a hardware relay, a safety‑rated STO/SOS function, or a software interlock, and how it is energized/de‑energized.
- “Redundant speed‑limiting” should specify the independent channel(s) implementing the limit (e.g., drive‑integrated safety function, separate PLC, or on‑drive configuration), interfaces, and whether redundancy is diversity‑based.
- “Self‑calibration” must identify what parameters are calibrated (e.g., torque constant Kt, viscous/Coulomb friction, gravity terms) and the model equations and conditions triggering and halting calibration.
- Define sampling rates, synchronization, sensor placements (IMUs on adjacent links across the target joint), filtering, and time alignment sufficient to reproduce the effects.
- Avoid pure functional claiming; if using means‑plus‑function in the U.S., provide corresponding structure/algorithm in the specification.
VI. Sufficiency of Disclosure/Written Description (35 U.S.C. § 112(a); EPC Art. 83)
To satisfy enablement and support, the specification should disclose:
- The dynamic model used for calibration, including how IMU measurements (angular rate/acceleration) and motor current map to joint torque and parameter updates; treatment of gravity/friction; and any identification method (e.g., EKF formulation with state vector, process/measurement noise models, or recursive least squares with forgetting).
- Dual‑IMU calibration (bias, scale, alignment), sensor placement tolerances, and synchronization strategy; mitigation of magnetic disturbances if magnetometers are used.
- Observability/identifiability conditions and gating logic; thresholds and statistical tests (e.g., residual variance windows) to prevent mis‑calibration.
- Fault detection logic, thresholds, timers, hysteresis; diagnostic coverage targets; and the reaction time of the circuit‑breaker and independent speed‑limit path.
- Architecture diagram indicating independence of the safety channel and its interfaces to actuators.
- Empirical or simulated performance demonstrating reduced drift, improved torque estimation accuracy, maintenance interval impacts, and false‑trip rates.
VII. Industrial Applicability (EPC Art. 57)
The invention is clearly susceptible of industrial application in warehouse robotics and AGVs/AMRs with articulated manipulators.
VIII. Unity of Invention
Unity is likely satisfied if the safety mechanisms are functionally tied to, and triggered by, the calibration residuals as part of a single general inventive concept. If claimed as separate, untethered features (a calibrator and an unrelated safety limiter), a lack of unity objection could arise.
IX. Claim Drafting and Amendment Recommendations
To fortify novelty/non‑obviousness and eligibility:
- Recite the dual‑IMU configuration across the joint with explicit fusion and synchronization steps that yield a relative joint dynamic estimate used to update torque model parameters Kt and friction terms online.
- Include observability gating: calibration proceeds only when residual statistics and excitation metrics meet defined thresholds over a time window; otherwise frozen.
- Define the residual R(t) between IMU‑derived joint torque and current‑derived electromagnetic torque, with bounded noise models and filters; use R(t) both as an estimator input and as a safety diagnostic.
- Specify the anomaly circuit‑breaker: independent actuation path, reaction time Treact, and fail‑safe behavior; define conditions for auto‑recovery versus latched stop.
- Define the redundant speed‑limit path implemented independently of the main controller (e.g., drive‑integrated safety function or separate safety PLC), triggered on consensus logic over residuals and sensor health checks.
- Provide quantitative performance targets (e.g., maximum allowable torque estimation error, diagnostic coverage, permissible speed envelope in degraded mode).
- Add dependent claims on: IMU placement tolerances; calibration algorithm (EKF/RLS); handling of payload variation; friction model; synchronization method; residual thresholds/hysteresis; compliance with specific safety integrity levels (without over‑promising).
X. Conclusion
- Eligibility: Likely patent‑eligible in both U.S. and Europe when claimed as a concrete control and safety system for a physical robot.
- Novelty: The individual elements are broadly known. Novelty may be sustained if the claims recite a specific dual‑IMU/self‑calibration integration with residual‑driven diagnostics and concretely defined safety responses and independence of the speed‑limiting channel.
- Inventive step: At risk under § 103/Art. 56 if presented as a straightforward combination of known sensors and safety practices. Non‑obviousness is supportable where the application demonstrates and claims a synergistic architecture addressing time‑alignment, bias/drift, observability‑gated calibration, and a residual‑driven, independent, redundant safety path achieving measurable improvements in accuracy, reliability, and maintenance efficiency.
- Formalities: Clarify terminology; disclose sufficient algorithmic and architectural detail; provide data supporting asserted performance advantages.
A prior‑art search focusing on: (i) IMU‑aided joint torque estimation and calibration; (ii) dual‑IMU arrangements on articulated links; and (iii) safety architectures with independent speed limiting for warehouse robots, is recommended to tailor the distinguishing features and finalize amendments.
