Analytics

Fleetera Analytics lets you define KPIs — tenant-defined calculations computed continuously by the Fleetera engine over your live telemetry. You describe the calculation once, target it at a class of assets, and the platform computes and stores a result series for every matching asset — no external BI tool, no SQL, no per-asset copy-paste.

You find KPIs under Analytics → KPIs in the sidebar. The catalog lists every definition with its cadence, the number of assets it currently applies to, its engine health, and an enable toggle.

How it works

KPI definition                     Fleetera engine                 Results
┌───────────────────┐   applies   ┌──────────────────┐   writes   ┌──────────────────┐
│  Calculation spec │ ──────────► │  Computes every  │ ─────────► │  One value per   │
│  (operator DAG)   │  to every   │  schedule window │            │  asset per       │
│  + population     │  matching   │  per asset       │            │  window, stored  │
│  + cadence        │  asset      │                  │            │  as a KPI series │
└───────────────────┘             └──────────────────┘            └──────────────────┘

A definition binds three things together:

• What to compute — a calculation spec built from a curated library of operators (aggregations, counter deltas, integration, resampling, classification, cycle detection, arithmetic).
• Where to compute it — a population selector (asset categories, optionally narrowed to a site). One definition applies to every matching asset, including assets added later.
• How often — a cadence from one minute to daily.

The engine evaluates the spec once per schedule window for each matched asset and writes one result point per asset per window. Results are first-class data: they are stored per asset alongside your telemetry, carry quality and input-completeness metadata, and record which definition revision produced them. Cycle-detection KPIs additionally record one entry per detected cycle.

Key concepts

• KPI definition — the tenant-level object holding the calculation spec, population, output declaration, and cadence.
• KPI key — the machine-readable output key (lowercase snake_case, for example avg_active_power). Unique within your organization and immutable after creation.
• KPI variables — enabling a definition creates a variable of kind kpi on every matched asset. This is where the computed series lives, alongside the asset's raw and derived variables.
• Revisions — every save of a definition creates a new immutable revision recording who authored it and when. The definition detail page lists the full revision history, so you always know which version of the calculation produced which values.
• Draft-first lifecycle — new definitions are always saved disabled. You preview real values first, then enable as an explicit second step. Disabling stops computation at the next tick; already-computed values are retained.
• Quality, not gaps — every result point records how complete its input data was. Points computed from incomplete input are written with quality uncertain rather than dropped, and windows where the calculation is not computable (for example a division by zero) write no point at all — the engine never fabricates a value.

Cadences

Each result point is timestamped at the start of its window. A daily definition therefore produces one value per asset per calendar day — no additional aggregation step is needed.

A site's timezone defaults to UTC, so daily windows are UTC days until you set one on the site. Changing it moves every day boundary, so the stored daily history for that KPI is recomputed on the new boundary.

Prerequisites

Before authoring KPIs, you need:

1. Assets organized into asset categories, so a population selector can target them
2. Variables on those assets with bindings delivering live data — a KPI can only compute over data that is flowing
3. For definitions with per-asset parameters (for example a nameplate capacity), the corresponding asset attributes set on each asset

Viewing the KPI catalog requires dashboard view access; creating, enabling, and installing KPIs requires dashboard management access. See user management for roles.

In this section