DATA METHODOLOGY
EV Charging Data Methodology
A practical explanation of how source records become a versioned, traceable EV charging dataset.
Reviewed 2026-09-28
Entity model
The canonical model separates a physical charging site from the equipment and commercial terms found there. A Station describes the place. An EVSE describes equipment that can serve one vehicle at a time. A Connector describes the physical interface and electrical characteristics. A Tariff describes pricing attached to a connector or charging service.
Source systems do not always expose every level. Published inventory fields therefore count records at their named normalization level; they should not be interpreted as a one-to-one ratio without checking completeness metadata.
- Station: location, address, access, parking, amenities and operator context
- EVSE: stable equipment identity, capabilities and maximum power
- Connector: standard, current type, voltage, amperage, power and status
- Tariff: currency, VAT treatment and energy, time, session or service components
Collection and normalization
Records are collected from approved public and partner sources. Source identifiers, timestamps and provenance are retained while field names, status codes, connector standards and pricing structures are mapped into the canonical model.
Potential duplicates are compared using identifiers, operator context, coordinates, address details and equipment attributes. A match is treated as a data-quality decision and remains traceable to its inputs.
Freshness and publication
The public snapshot is published on a daily cadence and exposes its own publication and dataset timestamps. Individual source observations may have different timestamps. Consumers should use field-level observation times for freshness-sensitive decisions instead of assuming every field changed at publication time.
Quality boundaries
Availability is an observed status, not a guarantee that a connector will remain free. Rated connector power is not measured grid load or delivered energy. Infrastructure data alone does not reveal utilization, sessions or revenue. Pricing and access conditions can change between observations.
Coverage totals describe records available in the current publication. They are not an official census and should be interpreted with the included timestamps and completeness indicators.
Standards and references
The canonical model can be projected into OCPI 2.3.0-compatible structures. OCPI is used as an interoperability target, while the canonical model remains the internal source of truth because upstream implementations and optional fields vary.
Inspect a real data sample
The public JSON sample uses the same canonical hierarchy described here.