What AFIR data reveals about charging reliability

What AFIR actually requires

The EU's Alternative Fuels Infrastructure Regulation (AFIR) does more than set deployment targets. It requires charge point operators to publish live status data to National Access Points (NAPs) — government-run feeds that operators are legally required to maintain. The data follows the DATEX II standard and is, in principle, open.

That changes the ground rules. Reliability figures no longer have to come from the operator's own dashboard, on the operator's own terms. They can be measured from data the operator is obliged to publish and cannot quietly edit.

The uptime problem AFIR exposes

AFIR mentions uptime — but uptime, as the industry reports it, is a self-reported number, and self-reported numbers consistently look good. Networks publish uptime figures of 95–99%, while independent field studies find that drivers complete a charge on the first attempt only around 71–73% of the time. That gap is documented across multiple studies, including UC Berkeley's 2023 field assessment, the 2024 ChargeX report, and the 2026 Stanford/NREL analysis.

The gap is structural, not a rounding error. A charger can report itself available at the moment a driver arrives, fail to start a session, and that failure never touches the uptime figure. From the operator's definition the charger was up. From the driver's experience it was broken. AFIR's open feeds make that contradiction measurable rather than debatable.

What the data shows today

Rather than take operators at their word, Symbioen reads the same government-mandated feeds and computes reliability independently. Across Q1 2026, that meant tracking 12739 stations and 78,673 charging points in Sweden and Norway — every status event sourced from AFIR-mandated feeds.

This is the gap that self-reported uptime hides. Networks report 95–95% availability, yet roughly 1 in 10 charging sessions fails to complete in Sweden, and closer to 1 in 7 in Norway. A charger can be marked available the moment a driver arrives and still refuse to start a session — and that failure never touches the availability figure. AFIR's open feeds make that contradiction measurable rather than debatable.

These are not figures an operator can opt out of or recalculate to look better. They are derived from the same status events AFIR requires every operator to publish. When the number drops, it is the network that moved — not the measurement. The full breakdown by operator is in the Q1 2026 reliability report.

Why most AFIR content stops short

Search for AFIR and uptime and you will find compliance guides, regulation summaries, and vendor explainers — almost all written from the operator's side of the table: what you must publish, by when, to avoid penalties.Almost nothing is written from the other side: given that this data must be published, what does it actually say about how chargers are performing?

That second question is the one Symbioen exists to answer. We do not sell compliance tooling. We read the mandated feeds and report what they reveal — for buyers, lenders, and operators who would rather see the network than be told about it.

Going beyond the public feed

AFIR-mandated status data is the foundation — it is open, independent, and cannot be edited after the fact. But status alone does not tell you whether a charger actually delivered energy. For due diligence, SLA audits, and dispute resolution, Symbioen can cross-verify the feed against actual session data, on request.

Two sources, layered on top of the public feed:

• Operator session logs — Transaction and meter-value data shared directly by the operator under NDA, reconciled against the same AFIR status events. This closes the loop between what the feed reports and what the charging hardware actually delivered.
• Fleet and mystery-shopper data — real charging attempts from commercial fleets and structured mystery-shopping runs. Where the feed says available but a verified driver session fails, that is ground truth no status feed can fabricate.

The result is a reliability view that holds up under scrutiny — useful when a number is about to drive an investment decision, a financing covenant, or a contractual dispute. This layer is offered on request, not as a public benchmark, precisely because it depends on data the counterparty has chosen to share.

Where this is useful

• Operators benchmarking their own network against the market on a measure they cannot influence after the fact.
• Investors and lenders verifying a charging asset in due diligence, against the claims made in the data room.
• Property hosts and municipalities holding the operator on a site accountable to an independent, AFIR-sourced number.
• Fleets routing around chargers that report as available but fail to deliver a session.

See the independent operator reliability leaderboard for how these numbers compare across networks.