Sentinel Confidence
A charger can appear healthy when evidence is sparse, delayed, stale, or internally inconsistent. Sentinel Confidence shows what supports the conclusion—and what does not.
Sentinel estate intelligence
Health
Understand what the available evidence says about operational condition.
State · change · trends
Observability
See the coverage, freshness and provenance behind the operational view.
Coverage · freshness · provenance
Evidence stays attached to findings, so teams can see what changed and why it matters.
The evidence story
Two chargers can report the same top-level status while exposing very different supporting evidence. One may provide recent measurements, detailed faults, and clear state transitions. The other may provide a single status message with no independent context.
A useful monitoring system should not make those situations look equally certain. Sentinel Confidence starts by separating what the evidence suggests from how much credible evidence is available.
Health and Observability
A health view is useful only when the evidence behind it is clear. Sentinel keeps the operational conclusion and the quality of its evidence separate.
Operational view
What does the available evidence suggest about this charger?
Health summarises what the available, relevant evidence suggests about an asset’s operational condition. A responsible health view should show when it was evaluated, link to the evidence behind it, and remain inconclusive when the evidence is insufficient.
Built from relevant state, event, measurement and trend evidence.
Evidence view
How much credible evidence is available to support that view?
Observability describes the coverage, freshness, provenance, and usefulness of the evidence available for evaluation. It makes gaps visible instead of allowing “no reported fault” to become “healthy.”
Built from coverage, freshness, provenance, usefulness and corroboration.
Health describes the conclusion. Observability describes the evidence available to support it.
Explainability
Sentinel attaches an evidence trace to material findings so the operator can see which inputs supported the view, which conflicted, which were missing or stale, and when the evaluation occurred.
Explainable findings
Material findings keep the operational context and evidence needed to support investigation.
Observability inputs
Observability considers whether evidence is available, recent, attributable and useful for the question being asked.
Are the relevant capabilities available for the question being asked?
Is the evidence recent enough to support the displayed conclusion?
Can Sentinel identify the source, integration path, mapping, units, and applicable context?
Does the signal carry enough meaning, timing, and quality to contribute to the evaluation?
Is there sufficiently independent evidence to support or challenge the same conclusion?
Are missing, stale, unsupported, or conflicting inputs represented honestly?
Third-party assurance
These levels describe verified evidence depth for third-party integrations. They do not rate electrical safety, product quality or standards compliance.
01
Baseline operational evidence is available through a verified integration, with known limitations made visible.
02
Additional verified signals or context extend the baseline view.
03
A deeper supported integration provides documented mappings, provenance, and a broader evidence set.
04
A verified evidence interface supports the Sentinel Confidence framework.
FAQ
Sentinel Confidence shows both what operational evidence suggests and how much credible evidence supports that view.
Health describes the operational conclusion supported by available evidence. Observability describes the coverage, freshness, provenance, and usefulness of the evidence available to support that conclusion.
No. It describes evidence available to a monitoring system. It does not certify electrical safety, replace device protection, or approve a charger for use.
The charger may be operating normally, but Sentinel should not present a strong health conclusion when evidence is insufficient. Low observability describes uncertainty in the view, not proof of a fault.
No. Signal availability varies by OCPP version, implementation, model, firmware, configuration, and integration. Capabilities must be verified.
Sentinel treats missing and stale evidence as explicit conditions so uncertainty remains visible in the operational view.
No. PlugStream Native is a separate native-device category, not a fifth rung in the third-party assurance ladder.
Start with the chargers, integrations, and signals already in your estate. We will help you map what is visible, where evidence is limited, and which questions need a better data source.
