Digital Commons Network^™

Visual Movement Indicator For Head-Worn Displays, Chia Wei Kao

Defensive Publications Series

Collisions between users of head-worn displays and individuals in the immediate environment often occur because the physical movement of the user is unpredictable to bystanders. This disclosure describes a head-worn display equipped with external visual indicators and an integrated orientation sensor, such as a gyroscope. The sensor detects changes in head position or user leaning, and specific patterns of light are emitted from the visual indicators based on the detected direction. For example, forward, backward, or lateral movements trigger corresponding indicators on the front, rear, and/or sides of the device chassis. By providing real-time visual signals of a user's intended …

Just-In-Time Adaptive Interventions For Wellness Management Via Context-Aware Wearable Devices, Cecilia Abadie, Aveek Purohit, Pinal Bavishi, Theo Guidroz, Ayush Jain, Ines Mezerreg

Defensive Publications Series

ABSTRACT

Manual logging for health and wellness often results in low user engagement due to the high effort and friction involved for data entry. Additionally, ongoing activity observation using high-fidelity sensors on wearable devices is constrained by finite battery life and computational resources. To address these challenges, a unified architecture is described that facilitates just-in-time adaptive interventions.

A cascaded sensing layer is utilized to reduce power consumption. A low-power sensor screens for motion signatures indicative of potential events, such as hand-to-mouth gestures, on a persistent basis. Upon detection, a high-power sensor is triggered to capture rich contextual data. This data …

Retrieving Personalized Content Using Ai Agents And User Profile Information, Benjamin Azose, Akancha Gupta

Defensive Publications Series

Agentic workflows can automate many user journeys. However, when a user journey involves choosing between different options, e.g., dietary preferences, the AI agent may be inadequate in automating the user journey, requiring manual actions and/or user inputs to make a choice. This disclosure describes techniques that automate agentic workflows by permitting an AI agent to access user preferences from a user profile prior to performing a corresponding task. With user permission, user preferences are obtained via explicit input, user browsing history, transactions, etc. and are stored in a user profile that is accessible to AI agents. When summoned to perform …

The Gossett Protocol: A 120-Point Master Engineering Standard For Sovereign Neural-Interface Bypasses, Jon S. Gossett

Defensive Publications Series

This disclosure establishes a comprehensive 120-point industrial engineering standard for a semi-invasive digital spinal bypass and neural communication bridge. The protocol prioritizes "Human Sovereignty" through a hardware-locked, 16-bit discrete metadata standard, ensuring all motor intent is processed locally without cloud dependency. Key technical innovations include a 1.0mm non-penetrative cortical clearance zone, a medical-grade 850nm NIR optical bridge for transdermal signal transmission, and a 38.5°C thermal breach auto-kill safety wall. Version 6.2 introduces Category IX, detailing a "Camera-Assisted Grounding" process where external optical sensors are utilized exclusively for initial dictionary calibration and intent verification. This optical requirement is phased out once …

---

Go to article

Responsive User Interfaces Based On Task Criticality And User Context, Piyush Arora, Colby Hawker, Shantanu Pai, Ram Vivekananda, Wendy Yun, Jaime Sonoda

Defensive Publications Series

Responsive user interfaces enable dynamically adjusting user interfaces based on device-specific aspects such as screen size, aspect ratio, display resolution, etc. However, traditional responsive design fails to account for different types of constraints of a user and task criticality of the task being performed via the UI. Misalignment between the UI design, user context and task criticality can lead to user error. This disclosure describes techniques, implemented with user permission, for dynamically modifying the layout, information density, and/or interactive physics of a user interface based on a dual-factor analysis of user cognitive state and task criticality. The user's cognitive state …

Trust Amplification Exploits In Human To Ai Security Teams Abusing Overconfidence In Automated Recommendations, Pranav Bhatanagar Mr

Defensive Publications Series

Human–AI collaboration has become a central component of modern cybersecurity operations. Security analysts increasingly rely on automated systems for threat detection, risk assessment, and incident response. While this collaboration improves efficiency and scalability, it also introduces new vulnerabilities associated with human trust in machine generated recommendations. This paper introduces Trust Amplification Exploits, a novel class of adversarial strategies that manipulate and exploit overconfidence in AI-assisted security workflows. Rather than targeting algorithms directly, these attacks leverage repeated system accuracy, interface design, and organizational dependence to amplify human reliance on automated outputs. Over time, excessive trust reduces critical evaluation and increases susceptibility …

Interpretation Attacks: Exploiting How Ai Explains And Justifies Decisions (Turning Explainability Itself Into An Attack Surface), Pranav Bhatanagar Mr

Defensive Publications Series

Explainable Artificial Intelligence (XAI) systems are widely promoted as mechanisms for increasing transparency, trust, and accountability in automated decision-making. By providing human-readable explanations for model outputs, these systems are intended to support oversight, regulatory compliance, and informed human judgment. However, the growing reliance on automated explanations has created a previously overlooked security risk: the explanations themselves can be manipulated, exploited, and weaponized. This paper introduces Interpretation Attacks, a class of adversarial strategies that target how AI systems generate, present, and justify their decisions. Rather than manipulating model predictions directly, these attacks exploit the interpretability layer to influence human perception, distort …

Intent-Obfuscation Attacks: Evading Ai Security By Manipulating Meaning, Pranav Bhatanagar Mr

Defensive Publications Series

As artificial intelligence systems increasingly rely on intent-based security mechanisms to regulate user behavior, the nature of adversarial interaction is undergoing a subtle but significant shift. Rather than attempting to bypass safeguards through explicit policy violations or recognizable malicious patterns, attackers are increasingly manipulating meaning itself. This paper introduces Intent-Obfuscation Attacks, a class of semantic evasion techniques that exploit the gap between surface-level language compliance and the underlying objectives inferred by AI systems. Unlike traditional prompt injection or keyword-based attacks, intent-obfuscation operates through linguistic ambiguity, contextual indirection, and gradual semantic drift. Individual interactions appear benign and policy-compliant, yet when interpreted …

Ai Supply-Chain Pentesting: Attacking Models Through Training, Updates, And Dependencies (Hacking Ai Without Touching Production Systems.), Pranav Bhatanagar Mr

Defensive Publications Series

As artificial intelligence systems become embedded in critical decision-making workflows, security research has largely focused on defending models at runtime. Attention has centered on prompt injection, inference-time abuse, and output manipulation, implicitly assuming that models arrive in production as trustworthy artifacts. This assumption is increasingly fragile. Modern AI systems are assembled through complex and opaque supply chains involving datasets, pre-trained models, fine-tuning pipelines, dependency ecosystems, and automated update mechanisms. Each stage introduces opportunities for adversarial influence that remain largely unexplored by conventional security testing. This paper examines AI security from a supply-chain perspective, arguing that many impactful compromises occur upstream, …

Hallucination-Driven Exploits: Weaponizing Ai False Confidence In Cybersecurity Systems, Pranav Bhatanagar Mr

Defensive Publications Series

Artificial intelligence systems, particularly large language models, are now routinely used in operational cybersecurity environments. They assist analysts with alert triage, incident response, threat intelligence interpretation, and day-to-day decision-making in situations where information is incomplete and time pressure is constant. Much of the existing research on AI security has focused on direct attacks such as prompt injection, jailbreaks, or data poisoning. In contrast, this paper examines a quieter but increasingly consequential risk: the impact of confident hallucinations in security-critical contexts. When AI systems provide incorrect guidance with high confidence, the resulting decisions can materially affect defensive posture even when no …

Synchronous Multi-User Search Sessions With Group-Personalized Results, Paula Marques

Defensive Publications Series

Information retrieval systems designed for individual use may be inefficient for collaborative tasks, potentially requiring participants to search in isolation and manually share findings through disconnected channels, such as email or chat applications. A framework for synchronous, multi-user information retrieval sessions can synchronize participant actions in near real-time across multiple devices (e.g., smartphones, laptops, tablets, wearable devices). The system can allow a user to initiate a shared session and may generate a group-personalized set of results by synthesizing the profiles and contexts of the session participants. The framework may also provide an interactive collaboration layer with tools for shared bookmarking, …

Biometric-Based Dynamic Adjustment Of Vehicle Dynamics And Adas Parameters, Anonymous

Defensive Publications Series

The present disclosure describes a vehicle control system that utilizes occupant anxiety metrics—derived from in-vehicle sensors, mobile devices, and wearables—to modify the operational parameters of chassis control and Advanced Driver Assistance Systems (ADAS). When high levels of anxiety or stress are detected in the driver or passengers, the system shifts vehicle calibrations toward more conservative profiles. This includes reducing allowable wheel slip and yaw error, increasing suspension damping for stability, increasing ADAS following distances, and initiating earlier braking maneuvers. The system further enables ride-share integration, allowing a passenger’s wearable device to communicate anxiety levels to the vehicle to influence ride …

Technical Disclosure: Zero-Idle Asynchronous Hardware Architecture For Linear-Complexity State Space Models And Metabolic Gating, Samuel John Church

Defensive Publications Series

This disclosure describes a semiconductor architecture designed to eliminate dynamic idle

power consumption in Edge AI applications by utilizing clockless, event-driven logic to execute

linear-complexity State Space Models (SSMs). The system replaces the global synchronous

clock with local handshake protocols (Muller C-elements) implementing Pulsed Static CMOS

logic. This hardware substrate is coupled with a "Metabolic Gating" software router that

dynamically switches between low-precision "reflex" networks and high-precision "reasoning"

networks based on real-time energy availability and input "surprise" (prediction error). The

disclosed invention enables "Zero-Idle" operation, where power consumption scales strictly

linearly with input token arrival, achieving near-zero leakage during inter-token …

Link Token Provisioning, Yuexi Chen St

Defensive Publications Series

The present disclosure relates to the field of token management systems, in particular to mechanisms for enabling a token requestor to obtain a token from a token provider and to link a user credential to the token using a session code. One embodiment is related to a method. The method includes transmitting, by a token requestor computer to a token service computer, a token request message; receiving, by the token requestor computer from the token service computer, a session code; transmitting, by the token requestor computer to a user device, the session code, wherein the user device provides the session …

Power Optimization & Interference Management For Multicast Firmware Upgrade Over Lora, Praneetha Duggirala, Rajesh Mahapatra, Murali Bezawada

Defensive Publications Series

The disclosed invention relates to a power-optimized system and method for managing missing block transmissions and status reporting during multicast Firmware Upgrade Over the Air (FUOTA) in Long Range Wide Area Network (LoRaWAN) devices (e.g., smart meters). After firmware upgrade image transmission from a Head-End System (HES) to the devices in a Class C mode, the devices temporarily switch to Class A mode and transmit missing block information and status information frames at randomized intervals within a configurable window, minimizing network collisions and reducing power consumption. Devices then revert to Class C mode to receive multicast retransmissions and complete the …

Frequency-Traceability-Aware Timetransmitter Selection In G.8275.1 Full Timing Support Networks, Anonymous

Defensive Publications Series

Telecom packet timing networks based on the ITU-T G.8275.1 full timing support architecture typically use IEEE 1588 Precision Time Protocol (PTP) for time and phase distribution together with Synchronous Ethernet (SyncE) for frequency delivery, with upstream TimeTransmitter selection performed by the G.8275.1 alternate Best TimeTransmitter Clock Algorithm (A-BTCA). The existing A-BTCA prioritizes PTP dataset attributes such as clockClass, clockAccuracy, and variance, but does not consider the IEEE 1588 frequency traceability indication, even though frequency traceability can be critical in hybrid PTP+SyncE deployments. This disclosure identifies fault and holdover scenarios in which the current selection behavior can prefer a TimeTransmitter with …

Agentic Ai Framework For Semantic Transaction Data Compression And Retrieval, Palakh Shangle

Defensive Publications Series

The present disclosure relates to the field of Artificial Intelligence (AI), in particular to agentic AI framework for semantic transaction data compression and retrieval. The disclosed system comprises a plurality of AI agents, including an ingest agent, a selector agent, an encoder agent, a retriever agent and a governance agent, collectively configured to process large volumes of transaction records generated by electronic payment networks. In operation, transaction records are ingested, normalized, and analyzed to dynamically determine an appropriate compression policy based on storage efficiency, semantic fidelity and retrieval latency. Compressed transaction data is stored in an object storage system along …

Achieving Dynamic Urpf With Route Scale Optimizations, Anonymous

Defensive Publications Series

Systems and methods are disclosed for achieving dynamic Unicast Reverse Path Forwarding (uRPF) with route scale optimizations in an IP routing device, in which ingress traffic is sampled and analyzed to detect conditions indicative of source address spoofing or related attacks, and upon detecting such a condition, a control plane automatically enables uRPF at a selected scope including globally, per virtual routing and forwarding instance (VRF), and/or per interface, and subsequently disables uRPF once the attack condition subsides for at least a configured inactivity interval. Because enabling uRPF can reduce the maximum number of routes installable in a hardware forwarding …

Efficient Machine Learning And Serving Using Disjoint Hamiltonian Cycles, N/A

Defensive Publications Series

The technology described in this paper relates to the use of disjoint Hamiltonian cycles to support asymmetric communication patterns effectively while maintaining desirable latency control properties of direct-connect topologies. The technique used in this paper dynamically re-maps fixed disjoint dimensions of a direct-connect topology into a flexible set of disjoint Hamiltonian cycles that can trade network dimensions and the bandwidth-per-dimension to support asymmetric communication patterns. The mapping of the asymmetric network onto Hamiltonian cycles enables dividing the workload across Hamiltonian cycles of equal length.

Pipelined Execution Of Attention And Feed-Forward Network Operations On Heterogeneous Cores For Llm Decode Acceleration, N/A

Defensive Publications Series

The technology described in this paper relates to maximization of resource utilization for transformer based large language models (LLMs), specifically during an autoregressive decode phase. For the decode phase, each layer of a transformer model is primarily composed of two distinct computational blocks, an attention operator block and a feed-forward network (FFN) operator block. The attention operator block is overwhelmingly memory-bandwidth bound. In contrast, the FFN operator block is largely compute-bound, and the computations are dominated by large dense matrix multiplications. A chip architecture may be provided that contains at least two types of specialized co-located cores, a first core …

Packet Replication For Ai Training In Existing Data Centers, Zafar Ali, Pablo Camarillo, Clarence Filsfils, Ahmed Abdelsalam

Defensive Publications Series

Existing congestion management techniques for artificial intelligence (AI) data centers typically involve dedicated hardware for managing network congestion, which can be costly to implement, thereby requiring a large investment by network operators. Proposed herein is a technique to facilitate packet replication, which can make communications in an AI fabric resilient against congestion.

Automated Project Initiation And Management Using Generative Ai And Organizational Knowledge Retrieval, Katherine Brown Armstrong, Kevin Woods, Magdalena Flores

Defensive Publications Series

Significant administrative overhead is often encountered in operational environments due to fragmented, manual processes for tracking diverse work types and managing project lifecycles. Reliance on individual effort across disparate applications leads to inconsistent execution and limited visibility into workflows. To address these limitations, a method is disclosed for context-aware workflow automation utilizing generative artificial intelligence. Techniques are grounded in a curated organizational knowledge base containing standard operating procedures and best practices. Digital interactions are analyzed to derive project context, triggering automated cross-application actions such as drafting project artifacts, identifying stakeholders, and populating tracking tools. Administrative burdens are reduced by automating …

Testing Calibrated Delay Circuit Propagation Delay With A Phase Frequency Detector, Karthikeyan Subramanian, Kasi Chunduri, Mukesh Goyal, Dhivyabharathi Vethanayagam, Gurumoorthy, Durgesh Gupta

Defensive Publications Series

Testing calibrated delay circuit (CDC) propagation delay at high double data rate (DDR) speeds is often limited by inaccuracies and increased test time in automated test equipment (ATE). Traditional characterization methods may be impacted by external hardware parasitics that degrade accuracy, as well as specialized tools that may increase both test time and overall cost.

This disclosure describes an on-chip analog design-for-test (DFT) method that utilizes a phase frequency detector (PFD) from an existing phase-locked loop (PLL) to measure CDC delay. The calibrated clock and reference clock are routed to the PFD to convert the delay into a duty cycle. …

Method For Smart Left Turn And Merging With Powertrain Aware, Anonymous

Defensive Publications Series

Drivers frequently encounter significant challenges when making left turns onto main roads or merging into traffic, primarily due to difficulties in accurately assessing gaps in oncoming traffic. A critical concern arises from the misperception of approaching vehicle speeds and distances, a problem greatly exacerbated by the increasing prevalence of Electric Vehicles (EVs) and other quick-accelerating vehicles.

EVs deliver maximum torque instantaneously, allowing many to accelerate from 0 to 60 mph in under 5.0 seconds. This unique capability results in different acceleration characteristics compared to Internal Combustion Engine (ICE) vehicles, with some studies noting that EVs may accelerate more quickly. As …

Uart Virtualization, Debug And Trace In A Multi-Processor System, Mukul Mehta, Gopi Tummala, Prasad Avirneni

Defensive Publications Series

Accessing the UART (universal asynchronous receiver/transmitter) ports of a particular processor within a system-on-chip (SoC) generally requires exposing the UART pins of the processor at the die periphery. However, the die periphery has space only for a limited number of pins. This disclosure describes techniques that leverage universal serial bus type-C (USB-C) to communicate with processors over UART to enable frictionless SoC debugging. The techniques enable efficient use of SoC pins, rapid and parallel access to processor UARTs, and rapid root-cause analysis. On devices where USB is the only exposed interface, debugging can be facilitated by simply connecting the USB …

A System And Method For Real-Time Artificial Intelligence Driven Image Personalization For Digital Advertisements, Tarushi Dubey, Sumit Kumar, Pradeep Choube, Shivam Mohan, Vinay Sarda, Ravi Shanker Kumar Sinha

Defensive Publications Series

The present disclosure provides a method and a system for real-time artificial intelligence driven image personalization for digital advertisements. The method includes collecting the user data comprising user behavior data, and user transactions data and generating a user preference vector and a personalization sensitivity vector based on the user data. The method further includes determining a base advertising image comprising editable and locked components. The method includes determining personalized editable components based on the personalization sensitivity vector and generating a personalized advertising image based on the editing of the personalized editing components, user preference vectors and brand guidelines. The method …

System And Method For Preventing Insider Fraud Using Change-And-Revert Pattern Detection In Sensitive Banking Fields, Alok Roy

Defensive Publications Series

The present disclosure relates to a technique for preventing insider fraud using change-and-revert pattern detection in sensitive banking fields. The technique includes a monitoring framework configured to detect alterations to protected customer identity fields, such as mobile numbers and email addresses, recovery contact, card details, communication preference, or authentication identifiers. An audit trigger logs modification events, capturing details such as old and new values, timestamps, and user identifiers. A change data capture service continuously polls these logs to identify changes and activate cooling-off periods and transaction restrictions upon detection. The policy enforcement module invokes these restrictions, while the fraud pattern …

Fusing Fitness And Health Sensor Data From Multiple Sources, Justin Phillips, Daniel Roggen, Robert Harle

Defensive Publications Series

Estimates of cardiorespiratory fitness metrics (e.g., maximal oxygen uptake) from wearable electronic devices are often characterized by significant noise and variability. Inconsistencies arise due to algorithmic differences between manufacturers, sensor inaccuracies, and biological variance. The integration of data from multiple disparate sources lacks a standardized method for weighting measurements based on source reliability or handling late-arriving data. While the disclosed method is described in the context of cardiorespiratory fitness metrics such as VO2 max, it is equally applicable to any changing physiological variable derived from noisy or disparate sources, including but not limited to daily or weekly average blood pressure. …

Automated Detection And Identification Of Content Sources Using Invisible Watermarking, Anonymous

Defensive Publications Series

Systems, methods, and computer-readable media are described for automated detection and identification of content sources using invisible watermarking.

Techniques For Qr-Enabled Card Capture For Mobile Payments, Dileep R. Dominic

Defensive Publications Series

The present disclosure relates to a method for facilitating secure card-not-present (CNP) mobile payments through the use of a physical payment card embedded with an encrypted QR code. The QR code encodes essential cardholder data, including a primary account number (PAN) and expiry date. The method involves recognizing a designated input field on a mobile device while accessing a payment gateway interface. An interface module displays a camera icon adjacent to the input field, enabling a scanning procedure that utilizes a native camera application to read the QR code. This scanning process automatically populates the input field with the decoded …