Derivative works — US Patent 9729693

Here is the comprehensive "Defensive Disclosure" document for US patent 9,729,693.

Defensive Disclosure: Methods and Systems for Device State Determination

Publication Date: May 10, 2026
Subject Matter: Derivatives and expansions of the methods for determining the operational state of a sensor device, specifically whether it is worn by a user, correctly seated on a surface, or in a state of non-use, based on the teachings of US patent 9,729,693.

Derivative 1: Multi-Modal Sensing with Capacitive and Multi-Spectral Analysis

Axis: Material & Component Substitution
Enabling Description: This variation replaces a single-wavelength light emitter and photodiode with a multi-spectral optical sensor array and a co-located capacitive proximity sensor. The light emitter is a Vertical-Cavity Surface-Emitting Laser (VCSEL) array capable of emitting light at 850nm and 940nm, and the detector is a corresponding photodiode array. The capacitive sensor consists of two concentric copper traces integrated into the device's substrate. The worn-detection process is a multi-stage sequence:
1. Capacitive Proximity Check: The system first measures the capacitance between the traces. A significant change in capacitance above a pre-calibrated threshold indicates proximity to a dielectric material with high permittivity, such as human skin. This functions as a preliminary gating condition, replacing an ambient light check.
2. Multi-Wavelength Reflectance Ratio: Upon passing the capacitive check, the VCSEL array is activated. The system measures the reflectance intensity of both the 850nm and 940nm wavelengths. The ratio of reflected 940nm light to 850nm light is calculated. Human skin possesses a characteristic reflectance ratio in this spectrum. If the calculated ratio falls within a specific range (e.g., 0.9 to 1.3), it confirms the object is skin-like.
3. Signal Quality Metric: The standard signal quality metric calculation, derived from Photoplethysmography (PPG) analysis of the optical signal, proceeds to confirm a physiological signature.
  This composite method provides more robust "worn" detection that is less susceptible to ambient light interference and can better differentiate human skin from other dark, non-conductive materials.

Mermaid Diagram:

flowchart TD
    A[Start] --> B{Measure Capacitance};
    B --> C{Capacitance > Threshold?};
    C -- No --> D[State: Not Worn];
    C -- Yes --> E{Activate VCSEL Array (850nm & 940nm)};
    E --> F[Measure Reflectance];
    F --> G{Calculate Ratio (940nm/850nm)};
    G --> H{Is Ratio within Skin Range?};
    H -- No --> D;
    H -- Yes --> I{Calculate PPG Signal Quality Metric};
    I --> J{SQM > Threshold?};
    J -- No --> D;
    J -- Yes --> K[State: Worn];
end

Derivative 2: Worn Detection via Acoustic Impedance Plethysmography

Axis: Material & Component Substitution
Enabling Description: This disclosure describes a system where the optical sensor is replaced for the initial "worn" check with a miniature piezoelectric ultrasonic transducer operating in the 2-5 MHz range, designed to measure acoustic impedance. The method is as follows:
1. Acoustic Impedance Measurement: The system emits a short ultrasonic pulse from the transducer into the adjacent medium and analyzes the reflected echo.
2. Impedance Thresholding: The processor calculates the acoustic impedance of the medium from the echo's time-of-flight and amplitude. Human tissue has a characteristic acoustic impedance of approximately 1.5-1.6 MRayls. The system verifies that the measured impedance is significantly higher than that of air (e.g., > 1.0 MRayls) to confirm physical contact, replacing the optical voltage checks.
3. Signal Fluctuation Analysis (Acoustic Plethysmography): Following contact confirmation, the system polls at a higher frequency (e.g., 100 Hz) and analyzes subtle fluctuations in the echo's amplitude corresponding to changes in blood volume. This serves as the basis for the signal quality metric.
4. Worn Confirmation: If the signal quality metric derived from the acoustic signal exceeds a threshold, the device state is confirmed as "worn." This method is immune to optical interference from ambient light, tattoos, or skin tone variations.

Mermaid Diagram:

sequenceDiagram
    participant Processor
    participant Transducer
    participant UserSkin
    Processor->>Transducer: Send 2MHz Pulse
    Transducer->>UserSkin: Ping
    UserSkin-->>Transducer: Echo
    Transducer-->>Processor: Return Echo Data
    Processor->>Processor: Calculate Acoustic Impedance
    alt Impedance > 1.0 MRayls
        Processor->>Transducer: Begin 100Hz Polling
        loop Test Period
            Transducer->>UserSkin: Ping
            UserSkin-->>Transducer: Echo (with micro-variations)
        end
        Transducer-->>Processor: Stream Echo Amplitudes
        Processor->>Processor: Calculate Fluctuation SQM
        alt SQM > Threshold
            Processor->>Processor: Set State: Worn
        else
            Processor->>Processor: Set State: Not Worn
        end
    else
        Processor->>Processor: Set State: Not Worn
    end

Derivative 3: Temperature-Compensated Worn Detection for Extreme Environments

Axis: Operational Parameter Expansion
Enabling Description: This derivative specifies a "worn" detection system for use in extreme temperature environments, such as for personnel handling cryogenic liquids (-196°C) or working near industrial furnaces (200°C). The device body is a ceramic composite, and the optical sensor is a Silicon Carbide (SiC) based photodiode. The operational algorithm is modified for temperature compensation:
1. Dynamic Thresholding: The threshold voltages and the minimum signal quality metric are not fixed values. They are functions of temperature, retrieved from a look-up table (LUT) stored in non-volatile memory. An integrated thermistor provides real-time temperature data to the processor, which selects the appropriate threshold values.
2. Emitter Power Compensation: The drive current to the light emitter is dynamically adjusted based on the same temperature reading to maintain a constant photon output, counteracting the temperature-dependent efficiency curve of the emitter.
  This enables the device to reliably distinguish between being worn by a user and being placed on an inanimate hot or cold surface.

Mermaid Diagram:

stateDiagram-v2
    [*] --> Checking
    state Checking {
        direction LR
        [*] --> GetTemp
        GetTemp --> LoadThresholds: Read Thermistor
        LoadThresholds --> V1_Check: Load T_compensated thresholds from LUT
        V1_Check --> V2_Check: V1 < T1(temp)
        V1_Check --> NotWorn: V1 >= T1(temp)
        V2_Check --> SQM_Check: V2 > T2(temp)
        V2_Check --> NotWorn: V2 <= T2(temp)
        SQM_Check --> Worn: SQM > M1(temp)
        SQM_Check --> NotWorn: SQM <= M1(temp)
    }
    Worn --> Checking: on timer_event
    NotWorn --> Checking: on timer_event

Derivative 4: Aerospace Connector Seating Verification System

Axis: Cross-Domain Application
Enabling Description: This applies the core mechanism to verify the full mating of critical electrical or optical connectors in aerospace applications, such as avionics Line-Replaceable Units (LRUs). The female half of the connector is equipped with a micro-optical emitter/detector pair. The male half has a small, coated reflective surface that aligns with the sensor when fully mated.
1. Emitter Off (Stray Light Check): The system measures ambient light. A low reading passes, while a reading above a First Threshold indicates the connector is unmated in a lit environment, which may be logged as a maintenance fault.
2. Emitter On (Reflectivity Check): The emitter activates. A reflected light measurement greater than a Second Threshold confirms the presence of the specific mating surface.
3. Signal Quality (Electrical Continuity Check): The "signal quality metric" is an electrical measurement. The system performs a time-domain reflectometry (TDR) pulse on a data pin. A clean TDR response (metric > threshold) confirms a solid electrical connection.
  The final confirmed state is "Fully Mated," logged by the vehicle's Health and Usage Monitoring System (HUMS).

Mermaid Diagram:

graph LR
    subgraph Connector Mating Check
        direction LR
        A(Unmated) -- Begin Mating --> B{Ambient Light < T1?};
        B -- Yes --> C{Reflected Light > T2?};
        C -- Yes --> D{Perform TDR on Pin};
        D -- TDR Metric > M1 --> E(Fully Mated);
        B -- No --> F(Fault - High Ambient);
        C -- No --> G(Partially Mated);
        D -- TDR Metric <= M1 --> G;
    end

Derivative 5: Livestock Smart Collar Verification with GSR

Axis: Cross-Domain Application
Enabling Description: This system adapts "worn" detection for a smart collar for livestock to ensure it is on the animal and fitted correctly. The sensor module includes an optical emitter/detector and a galvanic skin response (GSR) sensor.
1. Optical Checks: The standard two-voltage optical checks are performed to verify proximity to a dark, skin-like surface, distinguishing the animal's neck from the ground.
2. Galvanic Skin Response (GSR) Baseline: After optical checks pass, the system measures skin conductance via the GSR sensor. A reading within the typical range for bovine skin confirms contact with a live animal.
3. Composite Signal Quality Metric: The signal quality metric is a composite value derived from the PPG sensor data (to detect a heartbeat) and a 3-axis accelerometer. The metric is high only if a low-frequency heartbeat is detected and the accelerometer signature matches the characteristic motion of the animal (e.g., grazing).
  A confirmed "worn" state validates collected health data and triggers an alert if the status changes.

Mermaid Diagram:

sequenceDiagram
    participant Collar
    participant Cow
    Collar->>Collar: Perform Optical Checks (V1<T1, V2>T2)
    alt Optical Checks Pass
        Collar->>Cow: Measure GSR
        Cow-->>Collar: Return Skin Conductance
        alt GSR in Bovine Range
            Collar->>Cow: Measure PPG and Accelerometer Data
            Cow-->>Collar: Return Biometric & Motion Data
            Collar->>Collar: Calculate Composite SQM (HR + Motion)
            alt Composite SQM > Threshold
                Collar->>Collar: Set State: Worn
            else
                Collar->>Collar: Set State: Not Worn (e.g., loose fit)
            end
        else
             Collar->>Collar: Set State: Not Worn (e.g., on fence post)
        end
    else
        Collar->>Collar: Set State: Not Worn (e.g., on ground)
    end

Derivative 6: AI-Based Multi-State Worn Status Classifier

Axis: Integration with Emerging Tech
Enabling Description: This derivative replaces fixed-threshold checks with a machine learning classifier (e.g., a Support Vector Machine or lightweight neural network) running on an edge AI processor. The system assembles a feature vector from a suite of low-power sensors: [v_optical_off, v_optical_on, capacitance, temperature, accel_std_dev]. This vector is input to a pre-trained ML model, which outputs a probability distribution across several states: [Worn_UserA, Worn_UserB, On_Table, In_Pocket, On_Charger]. The model is trained on a large dataset of sensor readings from these scenarios, enabling it to learn complex, non-linear relationships. A "worn" determination is made if the probability for any "Worn" state exceeds a confidence threshold (e.g., > 0.95), allowing for user-specific recognition and greater robustness than static rules.

Mermaid Diagram:

flowchart TD
    subgraph Data Collection
        A[Optical Sensor] --> E;
        B[Capacitive Sensor] --> E;
        C[Thermistor] --> E;
        D[Accelerometer] --> E;
    end
    subgraph Processing
        E[Create Feature Vector] --> F[ML Classifier (SVM/NN)];
        F --> G[Output State Probabilities];
    end
    subgraph Decision
        G --> H{P(Worn) > 0.95?};
        H -- Yes --> I[State: Worn];
        H -- No --> J[State: Not Worn];
    end

Derivative 7: Graded Confidence Score with Failsafe Low-Power Mode

Axis: The "Inverse" or Failure Mode
Enabling Description: This system calculates a continuous "worn confidence score" (0-100%) rather than a binary state. The score is a weighted average from multiple checks (e.g., optical check: 40%, normalized Signal Quality Metric: 50%, motion consistency: 10%). The system defines operational states based on this score:
- Confidence < 20% (Failsafe Mode): The device powers down high-drain sensors (GPS, etc.) and performs only a low-power "worn" check periodically (e.g., every 10 minutes).
- 20% <= Confidence < 80% (Limited Functionality): The device collects sensor data but flags it as "low confidence."
- Confidence >= 80% (Full Functionality): The device operates normally.
  This provides a nuanced approach to data validity and drastically improves battery life when the device is not in use.

Mermaid Diagram:

stateDiagram-v2
    state "Failsafe (Low Power)" as LowPower
    state "Limited Functionality" as LimitedFunc
    state "Full Functionality" as FullFunc

    [*] --> LowPower: Initial State
    LowPower --> LimitedFunc: on check: Confidence > 20%
    LimitedFunc --> LowPower: on check: Confidence < 20%
    LimitedFunc --> FullFunc: on check: Confidence > 80%
    FullFunc --> LimitedFunc: on check: Confidence < 80%

Combination Prior Art Scenarios with Open-Source Standards

Combination with Bluetooth Heart Rate Profile: The "worn" detection method is combined with the standard Bluetooth GATT Heart Rate Service (HRS). The device only begins advertising the HRS and allowing connections after its "worn confidence score" (Derivative 7) exceeds 80%. If the score drops below this threshold, the device terminates the Bluetooth connection. This prevents client devices from maintaining a connection to a wearable that is not actively measuring data, conserving power for both devices. Open-Source Standard: Bluetooth SIG adopted specification for the Heart Rate Profile (GATT Service UUID 0x180D).
Combination with MQTT for IoT: The device's "worn" status is published as a retained message to an MQTT broker using the open-source MQTT protocol (ISO/IEC 20922). The topic is formatted as devices/{deviceId}/status/is_worn, with a payload of true, false, or a JSON object containing the confidence score. An external system running open-source software like Home Assistant can subscribe to this topic to trigger automations, such as adjusting lighting or thermostat settings based on whether a user is wearing a sleep-tracking or activity-monitoring device.
Combination with Android Sensor HAL: The entire "worn detection" logic is implemented as a new virtual sensor of type SENSOR_TYPE_WORN_DETECT within the Android Hardware Abstraction Layer (HAL). This provides a standardized interface for any application to receive "worn" or "not worn" events without needing to know the underlying hardware implementation (optical, capacitive, etc.). An application can register a listener to this sensor to automatically pause a workout log when the device is removed. Open-Source Standard: The Android Open Source Project (AOSP) specification for the Sensor Hardware Abstraction Layer.