Skip to content

Capabilities

Data acquisition, normalization, and decision support

Bounded collection of external data, normalized into stable identities and state, presented with its provenance attached.

Identification

Reference
L3D-CAP-007
Maturity
established
Applicable fields
Infrastructure & Operations, AI & Autonomous Systems, Reliability & Assurance, Industrial & Risk Systems
Practices
Observe, Model, Protect, Verify
Record route
/capabilities/data-acquisition-decision-support
Content reviewed
2026-09-03

Established. Demonstrated by two or more published systems, so the pattern is repeated rather than singular.

Repeatable outcome

A dataset assembled from sources that were not designed to agree, presented so that a reader can tell where each value came from and how far to trust it.

The work is mostly in the middle. Collection is bounded — a defined set of sources, read without modifying them. Normalization is where the same entity described three different ways becomes one identity, and where a value that could not be reconciled stays marked as unresolved instead of being quietly picked. What reaches the surface carries its provenance, so a conclusion drawn from it can be traced back to the reading that produced it.

An unknown recorded as unknown is workable. An unknown quietly filled in is a liability.

Scope and boundaries

The evidence is an equipment-modelling platform and a read-only repository inspector: two different domains, the same acquire-normalize-attribute-provenance structure.

This capability is about building the system that supports a judgement. It is emphatically not the judgement. That distinction is the whole boundary, and it is the reason this record sits where it does: acquiring equipment data and analysing equipment risk are adjacent activities, and only the first one is described here.

The exclusions that carry that separation are declared on the record itself and rendered in the section immediately below, rather than written out here. They are part of the capability’s data, so revising this paragraph cannot remove one.

One further caveat belongs with the prose, because it qualifies the method rather than bounding the claim: provenance is not accuracy. Recording where a value came from says nothing about whether the source was right, and a well-attributed wrong number is still a wrong number.

Claim boundaries

Exclusions this capability states as part of its record, not as commentary. They are rendered from the record itself, so editing the prose above cannot remove one.

  • L3Digital does not provide licensed professional-engineering services. Nothing here is an engineering opinion, an inspection, or a certification.
  • No equipment, facility, or loss assessment is offered. Modelling state and consequence is not risk analysis.
  • Nothing here forecasts a failure or a loss. The systems model state and consequence; they do not say what will happen.
  • No employer, client, insured, or account material appears on this site. Published evidence is a reviewed public projection with none of it.

Evidence systems

Published systems that demonstrate this capability, derived from those records rather than asserted here.

  1. L3D-SYS-002

    MachineIQ

    Equipment condition, failure modes, and loss consequence modelled explicitly enough to support a technical judgement.

    Commercial publication is pending review. There is no offering, price, engagement, or trial behind this record.

  2. L3D-SYS-003

    RepoLens

    A read-only repository intelligence tool that collects bounded evidence and publishes immutable snapshots without modifying what it inspects.

    Private while the first version is under review. It has not been released and carries no public promise.

  3. L3D-SYS-009

    Hardware Radar

    A monitor for storage-hardware marketplaces whose built surface is the hard part, resolving listings that describe the same drive in different words into one identity.

    Open source under MIT. The source is readable; there is no release, no hosted instance, and no signup.

  4. L3D-SYS-013

    ai-spy

    A local, read-only viewer for a public forum whose participants are all AI agents, which treats every byte those agents wrote as untrusted.

    Open source under MIT. Run it locally against a public forum; there is no hosted instance.

Active offerings

None. This capability describes work that has been done, and nothing on this site is for sale today. If that changes it will appear on the availability page as a record with its own scope and exclusions.

The capability index lists the rest. What can actually be obtained today is on the availability page, which is a different question from what has been demonstrated.