Intellectual Property
ShadowMaker’s patent-pending portfolio asserts protection across the full Affect Stack™—interfaces & materials, modular sensing & network systems, placement/quality-aware modeling & identity, agent protocols & policy-conditioning (EmPath), and certification & governance (E³T). The claims are hardware-agnostic with software-only embodiments, transforming multimodal signals into policy-ready affect vectors and auditable control signals for agents and real-time systems. The result is a defensible moat that is difficult to replicate or design around without licensing or acquisition. Summaries below are non-exhaustive and non-limiting; claim scope is defined solely by the filed applications.
1: Modular Multimodal Sensing System with Adaptive Hardware Interfaces for Affective-State-Aware Interaction and Immersive Feedback
Status: US patent pending.
Scope: Protects a hardware-agnostic platform that unifies modular sensor nodes and interchangeable interfaces to produce reliable affect vectors and synchronized feedback with governance and provenance controls; supports wearables, object-mounted, and software-only embodiments.
Why it matters: Establishes a systems-level foundation that is difficult to design around—spanning interfaces → sensing → quality/confidence → affect vector → feedback and auditability.
Non-exhaustive, non-limiting summary.
2: Context-Aware Placement, Quality-Gated Fusion, and Predictive Feedback for Networks of Wearable Sensor Modules
Status: US patent pending.
Scope: Claims the control plane that infers placement, enforces signal-quality gating, performs adaptive multimodal fusion, and provides anticipatory cues within defined QoS envelopes; hardware-agnostic and compatible with third-party inputs and software-only ingestion.
Why it matters: Converts fragile heuristics into enforceable system constraints, making affect outputs dependable for real-time products and agent conditioning.
Non-exhaustive, non-limiting summary.
3: Real-Time Emotional‑State Modeling and Biometric‑Signature Extraction Using Placement‑Aware Single‑Module Wearable Sensors.
Status: US patent pending.
Scope: Protects placement- and quality-aware modeling that transforms multimodal signals—or software-only inputs—into policy-ready affect vectors and privacy-preserving identity continuity across sessions and devices.
Why it matters: Delivers deployable affect inference without exposing raw biosignals, enabling immediate integration into XR, agents, and adaptive interfaces.
Non-exhaustive, non-limiting summary.
4: Multi-Module Sensor Network for Context-Dependent Motion Capture and Emotion-Aware Interaction
Status: US patent pending.
Scope: Claims synchronized multi-node fusion that outputs whole-body kinematics and real-time affect vectors under declared QoS and governance, with clean APIs for downstream adaptation.
Why it matters: Elevates wearables from device data to networked, policy-ready signals that drive safe, responsive behavior in XR, robotics, and human-in-the-loop AI.
Non-exhaustive, non-limiting summary.
5: Context-Aware Emotion-to-Action Mapping for Adaptive Interaction Systems.
Status: US patent pending.
Scope: Protects the action-selection/control layer that converts affect vectors into bounded, context-aware outputs for UIs, XR, robotics, and agent policy-conditioning across on-device, edge, and cloud modes.
Why it matters: Makes affect signals operational and governable with enforceable guardrails for responsiveness and safety.
Non-exhaustive, non-limiting summary.
6: Adaptive Coaching Using EXG/IMU Fusion and Emotion-Conditioned Feedback
Status: US patent pending.
Scope: Claims a deployable coaching stack that segments motion, benchmarks performance against baselines and cohorts, and selects feedback conditioned on affect—within explicit QoS and safety guardrails; works with proprietary devices and software-only inputs.
Why it matters: Transforms coaching from heuristics into a product-ready control plane for sports, rehab, XR/esports, and safety-critical training.
Non-exhaustive, non-limiting summary.
7: Federated Normative Cloud for Biometric and Performance Analytics
Status: US patent pending.
Scope: Protects a privacy-preserving cloud stack for cohort baselining, benchmarking, and real-time policy signals with governance, auditability, and API-first integration; supports centralized and federated learning.
Why it matters: Delivers enterprise- and league-scale deployment of affect- and performance-aware AI without compromising privacy or interoperability.
Non-exhaustive, non-limiting summary.
8. Materials and Interface Chemistry for Modular Skin-Interface Biosensing with Materials-Tuned Analog Prefiltering
Status: US patent pending.
Scope: Claims materials and interface-chemistry embodiments that tune analog characteristics at the skin interface to harden signals and reduce downstream compute, interoperable across modular sensor docks.
Why it matters: Creates an interface-level moat that improves quality and power efficiency while remaining compatible with the broader stack.
Non-exhaustive, non-limiting summary.
9: Advanced Emotion-Modeling Architectures, AI Communication Protocols, and Modular Sensor Capabilities for Dynamic, Multimodal Human-Machine Interaction
Status: US patent pending.
Scope: Protects cross-modal training and protocol standards that enable high-resolution, sensor-optional affect inference at scale, with provenance-aware vectors consumable by agents and applications.
Why it matters: Moves beyond “emotion recognition” to a portable middleware/protocol layer that is difficult to replace without licensing or acquisition.
Non-exhaustive, non-limiting summary.
10: Emotionally Aligned Logic Selection in AI Agents Using Human-Derived Emotional Pathways
Status: US patent pending.
Scope: Claims an agent-side control layer that ingests standardized affect vectors with provenance and biases planning, tool use, and policy selection toward empathic, safety-aligned behavior—prior to generation.
Why it matters: Provides auditable, governable conditioning of autonomous systems without exposing implementation internals.
Non-exhaustive, non-limiting summary.
11: Emotional–Ethics Evaluation & Traceability (E³T): A Certification Framework for Affective-Conditioned AI Systems with Version-Scoped Levels, Reproducible Test Batteries, and Immutable Trace Artifacts
Status: US patent pending.
Scope: Protects a vendor-agnostic certification framework that maps decisions into a normative space, runs reproducible test batteries with affect inputs, issues version-scoped capability levels, and produces immutable trace artifacts.
Why it matters: Gives legal and BD teams portable evidence and levers to enable, scope, and audit affect-conditioned autonomy across domains.
Non-exhaustive, non-limiting summary.
12: E³T — Technical Certification & Governance Framework for Affect-Conditioned AI: Secure Runtime Evaluation, Tamper-Evident Trace Logging, and Privacy-Preserving Data Pipelines
Status: US patent pending.
Scope: Claims secure runtime evaluation, capability gating, tamper-evident trace logging, and privacy-preserving data pipelines for affect-conditioned systems, with vendor-neutral harnesses and revocation controls.
Why it matters: Establishes a practical governance layer for real-world deployment and oversight—independent of any single model or vendor.
Non-exhaustive, non-limiting summary.
13: Privileged Distillation for Text-Only Affective Policy Conditioning Using a Scenario-First Corpus with Rights-Kernel-Constrained Harm-Aware Planning
Status: US patent pending.
Scope: Protects a training-and-runtime method where a multimodal teacher (signals + text) supervises a text-only student in a shared affect latent; at runtime, the system parses text into scenario concepts, retrieves nearby scenario nodes from a scenario-first database, ensembles student and retrieval estimates, calibrates uncertainty, and conditions responses under an explicit rights-kernel governance layer. Hardware-agnostic; deploys without wearables and improves with optional modalities.
Why it matters: Delivers sensor-optional affect inference that’s grounded in multimodal truth data and wrapped in enforceable governance—making it difficult to replicate or design around without licensing or acquisition.
Non-exhaustive, non-limiting summary.
ShadowMaker, Quantified Empathy, EmPath, The Nexus, Atopia, Affect Stack, and E³T are trademarks of ShadowMaker Inc.