Lineage-Recorded Provenance

Mechanism

Each derived observation produced through any derivation function carries a derivation-lineage record. The record comprises references to each input observation, an identifier of the derivation function applied, a governance-policy version under which the derivation function is configured, a derivation timestamp, and a cryptographic integrity attestation over the derivation-lineage record. The record is intrinsic to every derived observation rather than being an optional metadata annotation. Within the five-property governance chain, every observation, evaluation, and action is linked through deterministic lineage across the architecture.

The derivation-lineage graph is a directed acyclic graph spanning derivation steps across any number of consuming agents, across any number of derivation functions, and across any number of intermediate transformations. A derivation-chain traverser supports bidirectional traversal: forward traversal from an input observation produces the set of derived observations transitively depending on that input, supporting impact-analysis queries; backward traversal from a derived observation produces the input observations and transitive inputs from which it was produced, supporting provenance-analysis queries. The lineage links are intrinsic to derived observations rather than embedded copies of upstream records.

Tamper-evidence is structural. The derivation-lineage record is cryptographically attested, producing tamper-evidence for the derivation path. Because the record commits to the references to its input observations, the derivation function identifier, the governance-policy version, and the derivation timestamp, modification of a recorded component is detectable through the integrity attestation. Every derived observation and every downstream decision has a reconstructible path back to the originating sensor observations, across any number of intermediate derivation agents, with every transformation identified and the governance-policy version recorded.

Operating Parameters

The derivation-lineage record supports regulated-domain deployment, where a regulated domain such as healthcare, defense, financial services, legal services, critical infrastructure, aviation safety, nuclear safety, or pharmaceutical manufacturing requires that every derived decision be auditable back to its source evidence. The derivation-lineage record provides the structural element satisfying this requirement: every derived observation and every downstream decision has a reconstructible path back to the originating sensor observations, with every transformation identified and the governance-policy version recorded.

A temporal-reconstruction mechanism supports reconstruction of the view state at any prior time for forensic, audit, regulatory, and legal purposes. It accepts a target time and target query scope, retrieves all observations admissible at the target time within the scope, applies the admissibility rules in force at the target time per the governance-policy-version lineage, applies the aggregation functions in force at the target time, and produces a reconstructed view consistent with the filters in force at the target time. A reconstruction-lineage recorder records each reconstruction request, its target parameters, and the reconstruction output.

Temporal reconstruction is governance-chain-preserving: the reconstructed view is itself a governed observation, with lineage linking it to the contributing historical observations and the governance policies in force at the target time, admissible as evidence in regulatory, legal, and governance-enforcement proceedings. Supported audit and forensic patterns include incident investigation, regulatory-compliance audit, attribution-audit, consent-compliance audit, performance audit, litigation-support reconstruction, policy-evaluation reconstruction, and any governance-policy-defined reconstruction pattern.

Alternative Embodiments

A derived observation's effective evidential weight is computed from its input observations' effective evidential weights through a governance-policy-defined derivation-weighting function, selectable from any of a plurality of forms. These include a minimum form, in which the derived weight is the minimum of the input weights; a product form, in which the derived weight is the product of the input weights; a weighted-average form; an uncertainty-aggregation form under a governance-policy-defined uncertainty model; a Dempster-Shafer combination form; a Bayesian combination form; a learned combination form; a composite form; or any equivalent weighting form producing a governance-chain-preserving derived weight.

Derivation functions within the scope of the lineage mechanism include multi-source sensor fusion, cross-domain observation escalation, integrity conflict resolution producing a reconciled observation from a conflict set, dispositional derivation, the forecasting kernels, capability-envelope composition, cascade propagation, discovery aggregation, observation routing, any governance-policy-defined derivation function, or any combination of the foregoing. Each such derivation function produces derived observations that are themselves governance-chain-preserving governed observations with their own derivation-lineage records, and derivation functions are governance-policy-configurable, admitting new derivation functions through governance-policy update.

The lineage construct recurs across the architecture under operation-specific recorders that preserve the common governance-chain-preserving form. Governed actuation carries a lineage record of the executed actuation, the observed effects, the verification outcome, and the actuation-state broadcast. Settlement and dispute operations record each dispute initiation, admissibility determination, and resolution outcome in the governance-chain lineage. The common lineage discipline lets a single audit traversal recover the contributing history regardless of which operation produced the record.

Composition

Lineage-recorded provenance composes with the other four properties of the governance chain through recursive chain closure. The five-property chain is recursive: observations generated at any primitive's output feed into the chain, and actuations emitted from the chain pass through every primitive's governance. Dispositional observations enter the chain as authority-credentialed observations, forecast observations enter with evidential weight from forecasting-kernel track records, discovery query results enter with per-observation lineage, cascade propagation observations enter with topology and propagation-function provenance, and settlement records enter with matched-pair or N-party attestation. Every primitive closes into the same five-property governance chain.

Cross-mesh reconciliation preserves lineage continuity across mesh boundaries, enabling multiple independent governed-mesh deployments to interoperate and reconcile divergent observation histories upon interconnection without requiring a prior shared authority or consensus protocol. Dispute mechanism integration is supported because each dispute initiation, admissibility determination, and resolution outcome is recorded in the governance-chain lineage, so the dispute history is recoverable by the same traversal that recovers the operation history. Because every observation, evaluation, and action is linked through deterministic lineage, a single audit traversal recovers the contributing decision history across the architecture.

Prior-Art Distinction

The derived-observation lineage mechanism is structurally distinguished from prior sensor-fusion output formats and prior data-pipeline output formats in that the derivation-lineage record is intrinsic to every derived observation rather than being an optional metadata annotation; the record is cryptographically attested, producing tamper-evidence for the derivation path; the lineage graph supports bidirectional traversal across any number of derivation steps and across any number of independent consuming agents, rather than being limited to local single-agent derivation history; and derivation functions are governance-policy-configurable, admitting new derivation functions through governance-policy update rather than embedding fixed derivation logic. Prior blockchain append substrates, by contrast, cannot retract or correct and do not preserve consumption-view history.

The distinction is that the five properties operating together as a governed chain produce machine behavior structurally distinguishable from conventional sensor-actuator systems regardless of sensing modality, deployment domain, or actuation type. A system that senses, evaluates through a fixed algorithm, and actuates without authority credentials, evidential weighting, composite admissibility, and lineage provenance is a conventional sensor-actuator system. A system that implements the five properties as a unified chain implements the governed spatial mesh architecture. The chain treats lineage-recorded provenance as a structural property co-equal with authority-credentialed observation, evidential weighting, composite admissibility, and governed actuator execution, rather than as an auxiliary logging concern.

Disclosure Scope

The disclosure covers the derivation-lineage record and its schema, the bidirectional derivation-chain traverser, the derivation-weighting functions, the cryptographic integrity attestation and the tamper-evidence it produces for the derivation path, the temporal-reconstruction mechanism supporting forensic reconstruction of view state at any prior time, and the composition with the other four properties of the governance chain. The scope contemplates lineage evolution: derivation functions are governance-policy-configurable and new derivation functions are admitted through governance-policy update, without embedding fixed derivation logic. The scope further extends to regulated-domain deployment, where every derived decision must be auditable back to its source evidence, and to cross-mesh reconciliation preserving lineage continuity across mesh boundaries.

The disclosure as recorded in U.S. Provisional Application No. 64/049,409 further encompasses the derivation-lineage record that links each derived observation to its input observations and to the derivation function applied, with the governance-policy version and derivation timestamp recorded; the temporal-reconstruction mechanism applying the admissibility rules, aggregation functions, and filters in force at a target time per the governance-policy-version lineage; and the recursive five-property governance chain in which every primitive's output enters the chain and every actuation passes through every primitive's governance. Operations contemplated within scope include authority-credentialed observation, evidential weighting, composite admissibility, governed actuator execution, and lineage-recorded provenance across regulated and commercial deployments.