How the Symbioen Index works

Four principles

Independence. The index cannot be bought or influenced - and that is precisely where its value comes from. Public scores use National Access Point and infrastructure status feeds aligned with AFIR data categories; publication obligations and source coverage vary by market.

Driver experience first. The index is not designed to make operators look good. It is designed to measure whether chargers prove service in the public event stream. Every metric - from Availability to Charging Success Rate - is chosen because it catches part of the gap between what an operator reports and what a driver encounters at the charger.

Integrity. Every base metric is traceable to public charger events. EVSE and station scoring use the same core methodology across the market; aggregation is documented separately for each comparison unit so weighting and small-cohort adjustments remain explicit.

Regulatory alignment. Definitions of uptime, availability, and reliability are still settling across the industry. When consensus forms or new regulatory standards take hold, Symbioen updates the methodology accordingly. The methodology is continuously refined to better reflect real-world conditions - and stakeholders are kept informed of any changes as they happen.

The problem with Uptime

The standard metric in EV charging is uptime: the share of time a charger is not in a fault state. Operators publish it. Tenders require it. Investors use it in due diligence. The problem is that self-reported uptime often looks better than the charging experience. Independent field studies have found much lower first-attempt charge success than the 95-99% uptime figures commonly used in marketing, tenders, and data rooms.

The gap is structural. A charger can report Available at the moment a driver arrives, fail to start a session, and still look healthy in a simple uptime report. The charger was, by a narrow status definition, up. From the driver's perspective it was broken.

Where the data comes from

Symbioen uses public National Access Point and infrastructure status feeds aligned with AFIR data categories. The public index uses status-event history. Session logs, meter values, and vehicle-side telemetry can improve confidence, but they are not part of the public-data evidence unless supplied separately.

What the Symbioen Index measures

The current Symbioen Index is computed at EVSE level first, then aggregated to station, operator, municipality, county, and country levels.

The index combines five core signals. Availability is the share of observed time when the EVSE appears operational and reachable in the status feed. It is important, but it is not the whole score. A clean-looking charger still needs to prove recovery and charging success.

Signal Quality

Signal Quality measures whether the event stream itself can be trusted: duplicate timestamps, impossible ping-pong status patterns, and suppressed update storms all reduce confidence. A charger that is noisy in the feed is often noisy in operation.

Device Reliability

Device Reliability measures what the equipment does as hardware. It looks at failure frequency and short working blocks that suggest oscillation, restart loops, or unstable connectivity. One input is MTBF - mean time between failures: how long the EVSE typically runs before the next observed problem. Short transient reconnect noise is filtered so that brief blips do not incorrectly count as hardware failure.

Operational Responsiveness

Operational Responsiveness measures how quickly and credibly a charger recovers after problems. One input is MTTR - mean time to recovery: how long it takes the EVSE to return from an observed problem to a credible working state. It combines repair speed with post-failure recovery proof: a charger that returns to Available but never proves a charging session remains less credible than one that fails, recovers, and then delivers service.

Charging Success Rate

Charging Success Rate is inferred from status-history transitions into charging or in-use states. A detected session is considered successful when it lasts at least two minutes. A failed short session that is followed by a successful retry within thirty minutes receives partial credit. If no successful sessions are detected, this component contributes zero rather than redistributing its weight: an unproven charger should not score as if it has proved real service.

Availability

Availability measures the share of observed covered time when the EVSE appears operational and reachable in the status feed. A charger that is technically stable but routinely offline or unreachable is not serving drivers, so this component stays central.

Utilisation Rate

Utilisation Rate tracks the time a charger spends in active charging use. In practice, it is the strongest public-data confirmation that a charger really works: someone arrived, connected, and the charger entered an active charging state. Low utilisation alone does not prove a technical fault; it can also reflect demand, location, or route patterns.

Utilisation is reported differently from the score. The Symbioen Index averages only over rated chargers, because a charger the feed cannot judge should not be counted as a zero. Utilisation answers a different question - how much of the declared park is actually being used - so at operator and regional level it is measured against every station, not just the rated ones. A station the index cannot score still counts in the denominator as unused, which pulls a group's utilisation towards the share of its park that demonstrably carries load.

How the Symbioen Index is composed

The Symbioen Index brings together the components above into a single score, covering the major failure modes visible in public event data: bad signal quality, hardware instability, slow or unproven recovery, failed charging sessions, and outright unavailability.

The index is calculated at the level of the individual EVSE. The result is a single score per EVSE, ranging from 0 to 100.

A station score is the average of its eligible connector scores, weighted by how long each connector was actually observed - a connector seen for one hour counts less than one seen all week. The weakest connector's score and an estimated station reliability are published alongside, but only as diagnostics. Keeping the score a plain average means one persistently broken connector pulls the station down in proportion: it is neither hidden by a healthy neighbour nor able to drag the whole site to zero.

Operator and regional views build on these same station scores, with two rules. First, only rated stations count: a charger the index will not score is left out entirely - never counted as a zero - so a site the feed cannot see does not quietly drag its operator or region down. Second, each rated station is weighted by how much rated infrastructure sits behind it, so a large proven park counts for more than a single connector that happened to score well. Operator leaderboards add a Bayesian adjustment on top (below) to steady small samples. Each report states the aggregation it uses.

Cleaning the data before scoring

Raw status feeds from National Access Points arrive dirty. The same event turns up twice, a charger flips between states faster than any real charger changes, and a recovery lasts five seconds before the fault returns. Score that naively and you reward a charger for one Available signal sitting between two hour-long outages.

Signal Quality catches obvious data-stream issues before they contaminate the score. Duplicate timestamps, impossible status sequences, and repeated flip storms reduce the Signal Quality component. Short offline connectivity blips are filtered where appropriate so transient reconnect noise does not become a false hardware failure.

Broken Episodes are still meaningful. When a charger generates status changes at a rate that indicates a software, hardware, or connectivity problem rather than real state changes, the pipeline suppresses the storm and records that suppression as a scoring signal.

Fair operator rankings

An operator running a single charging point that had a good week should not sit at the top of a national leaderboard above operators running hundreds of stations. Small samples produce unreliable averages - a one-station operator is a coin flip, not a data point.

Operator leaderboards use Bayesian averaging to correct for this. Each operator's score is pulled towards the global mean in proportion to how small their sample is. An operator with one station gets pulled significantly; an operator with two hundred stations is barely affected.

Weighting is therefore not universal across every leaderboard. The Q2 market report uses Bayesian adjustment for network rankings and an unadjusted average for municipality rankings. When rank movement uses a baseline from an earlier Symbioen Index version, the report discloses that limitation alongside the ranking.

When a score is not published

Not every EVSE receives a score. Rating a charger responsibly takes enough recent signal history, and a station only qualifies once enough of its active EVSEs clear that bar themselves. Where the feed barely notices a charger, we would be inventing precision we do not have — so we return an ineligible result and say which test it failed. An honest gap beats a confident number.

One exception stops a broken charger from escaping its score by going quiet. The feed reports only changes, so a faulted connector announces its fault once and then has nothing left to say. Rather than letting it drop out, we read that last known fault as an ongoing outage and keep the connector rated at the floor. This holds only while other connectors at the site are still reporting - if the whole site goes silent, we treat it as a data gap, not a verdict.