Abstract

A home monitoring service receives no activity events for six hours. For nearly two of those hours the internet gateway was offline, although the local hub kept collecting data. A naive rule treats the silence as “no activity” and automatically sends an intrusive alert, turns on a camera or dispatches a robot. Not receiving an event is not the same as the event having been detectable and absent. The same failure applies to industrial monitoring, network operations and critical infrastructure, where “no alarm” can simply mean lost telemetry.

First public disclosure of the Mothership: PCT/IB2026/055615, filed 4 June 2026 and publicly disclosed 16 July 2026. Full text at Technical Disclosure Commons: https://www.tdcommons.org/dpubs_series/12542/ European public disclosure (Zenodo, 16 November 2026): “EUROPE’S NEXT AI SECURITY CHALLENGE: WHO CONTROLS THE FINAL ACTION? — From EU AI Act and NIS2 to 6G, the Digital Euro, Industrial AI and Critical Infrastructure.” https://zenodo.org/records/24274259

Based on Indian Provisional Filing: 202631136555, filed 15 November 2026, and disclosed with the Mothership (PCT/IB2026/055615) on 16 July 2026.

This patent-pending technical concept (Embodiment 12A) proposes negative-observation evidence finality. Two statements are kept apart: “the system received no event” and “every observation opportunity the contract requires was present and complete enough to support an absence claim.” Only the second can support a consequential action.

How it works

  1. An observation-coverage contract fixes the event class, region, time window, source classes, completeness requirements and the consequences the evidence may support.
  2. An expected-opportunity matrix records which source could have detected which event, where and when.
  3. A health vector for each observation path covers sensor function, calibration, field of view, delivery and buffering. A gap register keeps missing or invalid intervals instead of hiding them. A coverage epoch changes when topology, policy or evidence changes materially.
  4. An admissibility test returns supported, limited, unresolved or unacceptable evidence. A signed absence claim is not treated as universal physical proof.
  5. The real operation is reconstructed: exact notification recipient, information disclosed, camera or microphone enablement, machine movement and downstream consequences.
  6. Each consequence class has its own evidence requirement. Qualified absence may allow a narrow status statement, yet still fall short for surveillance or safety-critical actuation.
  7. A scoped authorization binds the accepted evidence, target, consequence and epoch. The defender-side sink (the notification dispatch, camera-enable interface or device controller) checks it before the effect happens. With missing coverage, a stale epoch, invalid consent or late events, the specific effect is refused. Only a separately authorized, narrowly worded monitoring-quality warning may pass.

Sensors, MQTT and event streams, device shadows, local gateways and safety functions stay in place. Per-sample cryptographic checks and full history reconstruction at the final gate are not needed. A privacy-disabled sensor counts as a gap, not as evidence of inactivity. Smoke, gas and other independently authorized safety functions stay operable.

Prospective industry relevance (AI-generated; see notice below). In each mapping the gate sits where the effect actually occurs.

  • Google (Google Home / Android Home APIs): an on-device or local-home validator builds source capability and health records from events and keeps outages in the gap register. An automation stays non-effective until a participating command dispatcher or target device enforces the decision. Direct cloud or alternate-credential paths need separate coverage.
  • Samsung (SmartThings): a coverage check separates “event did not arrive” from valid negative evidence before a sensitive scene or device command. At the SmartApp alone, protection reaches only managed commands.
  • Apple (Apple Home, Matter): a platform-controlled service keeps per-source report validity, privacy constraints and evidence epochs before absence-driven camera, lock or comfort actions. Accessory pairing alone does not show that conditions were observed.
  • Amazon Web Services (IoT Core, Device Shadow, IoT Rules): a coverage service reconciles shadow state, sensing capability and event gaps, and a separate authorization decision gates a consequential rule action or target device. A reported shadow property alone cannot certify that a sensor would have detected every relevant event.
  • Microsoft (Azure IoT Hub device twins): a trusted local or edge accumulator checks report completeness and source epochs. The twin is contextual evidence, not the sole authorization for physical actuation.
  • Apache Flink (event-time watermarks): a watermark is one delivery-completeness input, never proof of detectability, field of view, calibration or consent. This is a strong anticipated prior-art combination.
  • OpenAI (Responses API, Agents SDK) and Anthropic (Claude tool use): at the application tool dispatcher or broker, the actual recipient, camera request or physical target is reconstructed. Coverage and authority are verified independently, and an act-specific authorization is required at the defender’s sink. A model response or a tool’s “success” does not prove a sensing opportunity or completed actuation.
  • Critical infrastructure, RAN and industrial AI: “no alarm” can mean telemetry loss. Qualified evidence is maintained beforehand, and a bounded gate covers consequential configuration changes. Deterministic safety interlocks stay in place, and slow validation stays out of radio PHY and HARQ loops.

These mappings identify possible complementary integration points, not missing capabilities or deficiencies in the named products. Sensor fault diagnosis, heartbeats, event-time windows, stream watermarks, device shadows, policy engines and device interlocks are established technology and substantial prior art. Whether the specific cooperation claimed is novel or non-obvious is a patent-examination question and is not presumed. Negative evidence does not prove an occupant’s absence, a medical condition or a clear physical path. Practical assurances depend on calibrated detectability, correct completeness rules, enforceable control of the final effect and measured testing.

Where to read the full filing: search “Sangam Das” or PCT/IB2026/055615 (pctib202655615) on WIPO PATENTSCOPE (https://patentscope.wipo.int/), or read the PCT file at Technical Disclosure Commons: https://www.tdcommons.org/dpubs_series/12542/

Basis and notice: The underlying invention is human-originated and based on the Mothership application PCT/IB2026/055615, filed 4 June 2026 and publicly disclosed 16 July 2026. This abstract and its industry and product comparisons were prepared with AI-assisted tools. They may contain errors, omissions, outdated facts or speculation. Readers must cross-check the original public PCT disclosure and the companies’ own documentation before drawing any conclusion. Company and product references are explanatory only. They imply no adoption, endorsement, validation, collaboration, licensing or infringement, and no finding on patent validity or third-party rights.

Embodiment 12A of the 39-part series: this publication is the separately designated Embodiment 12A in a 39-part technical disclosure series, comprising 38 numbered embodiments and Embodiment 12A.

License: The attached PDF is submitted under CC BY-NC 4.0 International, irrespective of the license displayed by the submission site.

KEYWORDS

AI Safety; Negative Evidence; Absence of Evidence; Sensor Coverage Completeness; Sensor Health Monitoring; Telemetry Loss Detection; Observation Gap Detection; Event-Time Watermarks; Smart Home Security; Matter; IoT Security; Device Shadow; Digital Twin; Home Monitoring; Elder Care Technology; Camera Privacy; Industrial AI Safety; Critical Infrastructure Protection; AI-RAN; Agentic AI Security; AI Tool-Call Security; Consequence-Bound Authorization; Execution Finality; Finality Sink; Revocation; Privacy by Design; EU AI Act; NIS2; Zero Trust; Legacy System Compatibility

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Share

COinS