There is a number in EV fleet operations that most managers have never been shown, and it changes how you think about charging infrastructure entirely. Charging networks report uptime in the high nineties — and yet industry data indicates only around 71% of charging attempts succeed on the first try. Both figures describe the same equipment. Uptime measures whether the charger was online and reporting; it says nothing about whether the vehicle plugged into it actually finished a session. For a depot running tight schedules, that gap is the whole problem: a missed charge is not an inconvenience, it is a service disruption discovered at five in the morning when the first shift arrives. Closing it means treating the charger as a maintained asset with a fault history, not as a fixture that either has power or does not. Book a demo and bring a week of session data — we'll show you your own first-attempt success rate in Fleet Rabbit, which is almost certainly not the number on your uptime report.
INTEGRATIONS · EV CHARGING SESSIONS
EV Charger and Charging-Session Integration
Session data, charger fault codes and connectivity status connected to work orders, alerts and audit-ready records — so a failed charge becomes a maintenance event rather than a surprise at shift start.
98.7–99.9%
Reported network uptime
but
~71%
Of charge attempts succeed first time
Same chargers, both figures true. Uptime measures whether the unit was online and reporting — not whether the vehicle plugged into it completed a session.
Why the Gap Exists
Four reasons, and none of them show up in an availability percentage.
Online is not the same as working
A charger with a frozen screen, a failed payment or authorisation step, or a damaged connector is still reporting itself as available. It has power, it has connectivity, and it cannot charge a vehicle.
Sessions fail silently
A session that terminates part-way leaves a vehicle short of charge with nobody notified. The truck looks plugged in all night and departs at forty percent.
Vehicles get unplugged
By mistake, or because someone needed the bay. Entirely mundane, entirely common, and invisible unless the session record is being watched rather than assumed.
Failed attempts are often not stored at all
The most consequential of the four. Where a system logs only completed sessions, a repeatedly failing charger produces no record — so it never appears in a report, never generates a work order, and keeps failing until somebody complains loudly enough.
The framing that fixes it
A charger is a maintained asset with a fault history, a service schedule and a failure pattern — not a fixture like a transformer. Where a procurement or a monitoring setup treats it as the latter, the system delivered behaves accordingly: it tells you the unit has power and nothing else. Traditional maintenance platforms track vehicles; an electrified depot needs one that treats the charging infrastructure as equally critical.
What a Session Record Must Contain
Seven fields. The third is the one that separates a system you can improve from one you can only observe.
1Vehicle, charger and connector identityWhich unit plugged into which port, so both the vehicle history and the charger history receive the event rather than one of them.
2Start, end and energy deliveredThe basic session shape — and the input to cost allocation per vehicle, per route and per depot.
3Failed and interrupted attempts, storedComplete session data including attempts that did not succeed. Without historical log data covering failures, nobody can explain a past outage to stakeholders or quantify whether an operational fix actually worked.
4Charger error codesTransmitted from the unit and recognised by the receiving system. A code that arrives and is not understood is functionally the same as no code at all.
5Charger state at the timeActive, occupied, faulted or offline. Context that turns a failed session into a diagnosis rather than an incident report.
6Departure readiness against requirementState of charge reached versus what the next route needs. The only version of "charged" that matters operationally.
7Site load at the time of the sessionBecause a session that failed during a demand peak is a load management problem, not a hardware one — and the two need opposite responses.
Point three deserves emphasis because it is usually a contractual matter rather than a technical one. Complete session storage including failed attempts, error code transmission and data ownership should be written into the agreement rather than assumed — several published reviews of current procurement note these requirements missing from live specifications. Ask for them before signing, because retrofitting a data right is considerably harder than requesting one.
71%
first-attempt success, industry-wide
What is yours? Most depots have never calculated it.
Bring a week of session logs to a 30-minute call. We'll load them into Fleet Rabbit, calculate first-attempt success by charger and by bay, and show which specific units are producing your failures. In most depots the answer is a small number of ports rather than a general reliability problem — and that is a repair list, not a replacement programme.
Treat the Charger Like Any Other Asset
Everything you already do for a vehicle applies here, and almost nobody does it.
A fault history per unit
Every error code, failed session and offline period recorded against the charger and the specific connector. Repeated connector errors are a pattern long before they are an outage, and they only exist as a pattern if somebody kept the record.
Work orders raised automatically
A fault threshold crossed should generate an assigned task with the error history attached — not a notification. The same discipline you apply to a check engine light applies to a charging bay.
Preventive maintenance schedules
Connectors, cables, cooling and enclosures wear. Scheduled attention on a charging asset is as legitimate as it is on a vehicle, and considerably cheaper than the truck roll that follows neglecting it.
Reliability tracked by charger and by manufacturer
Which units fail most, and which brands fail most across your sites. That comparison is the single most useful input to the next procurement round, and it requires nothing more than having kept the fault records.
Why this scales badly if ignored
A single broken connector is a minor inconvenience across a handful of vehicles and a genuine crisis across fifty. Fleets add electric vehicles considerably faster than they add charging infrastructure, so the same number of ports serves more units each year — and the tolerance for a faulted bay drops accordingly. The failure that was survivable last year is the one that misses routes this year.
The Depot Problems Session Data Solves
Four operational issues that look like infrastructure shortages and are frequently scheduling problems.
← Swipe to see all columns →
| Symptom | What it usually is | What the data enables |
|---|---|---|
| Vehicles queuing for chargers | More vehicles than ports, but often also unstaggered sessions | Scheduling so one charger serves several vehicles across a long dwell |
| Unexpected demand charges | Everyone plugging in at once at end of shift | Load management holding total draw below the contracted maximum |
| Vehicles departing under-charged | A session that failed or was interrupted overnight | Alerting on failed sessions and unplugged vehicles before departure |
| Charging costs nobody can allocate | Energy recorded at the site rather than at the session | Energy analytics per vehicle, per route and per depot |
Two of these four are capacity problems and two are visibility problems, and they present identically — as "we need more chargers". Session data is what tells them apart, and getting that wrong is an expensive mistake, since grid capacity constraints can turn an infrastructure expansion into a delay measured in months rather than weeks.
Specify It Properly
Six requirements to put in writing before hardware or software is selected. Most current specifications are missing several of them.
01
Open protocol compliance, at the right version
Open Charge Point Protocol compliance is the non-negotiable specification for commercial fleet charging — 1.6J as a minimum, 2.0.1 preferred for new deployments, since the earlier version lacks the smart charging profile granularity needed for departure-based scheduling at scale.
02
Error code transmission and recognition
The charger must send vendor error codes over the protocol, and the receiving system must recognise them. Both halves, specified explicitly — one without the other produces logs nobody can act on.
03
Complete historical session storage
Including failed attempts, with a stated retention period. Without historical logs you cannot explain past failures to anyone or demonstrate that an improvement worked.
04
Data ownership, stated
Whose data it is and how you get it out. Contract this rather than assuming it, because the answer becomes important at exactly the moment relations with a vendor are least comfortable.
05
Real-time access to troubleshooting data
Not a monthly report. The point of the data is acting on it while the vehicle is still plugged in and the bay can still be reassigned.
06
Confirmation for your specific models
Protocol support varies in practice, so confirm the platform has actually been commissioned against your charger models rather than relying on a general compatibility claim.
One Dashboard for a Mixed Fleet
Most fleets electrifying today run both, and will for years. Splitting them across two systems creates the same problem twice.
One asset listElectric and combustion units on the same roster, with battery health on one and fuel consumption on the other rather than two separate operations.
One maintenance queueVehicle faults and charger faults in the same place, since a faulted bay disrupts routes exactly as effectively as a faulted truck.
One cost-per-mile comparisonEnergy cost per mile against fuel cost per mile, on the same basis. This is what the next procurement decision actually rests on.
One inspection and compliance recordDVIRs, defects and compliance history covering the whole fleet — the obligations do not change with the drivetrain.
Battery health alongside itPack degradation varies enormously between units even at similar ages, so state of health belongs on the asset record where charging behaviour and route assignment can be compared against it.
Turn a failed session into a work order, not a phone call at 5 AM
Fleet Rabbit connects to open-protocol chargers, records every session including failed attempts, raises work orders from charger fault codes with the error history attached, and keeps charging data on the same platform as vehicle maintenance, inspections and compliance — for electric and combustion units together.
What to Track
Six figures. The first is the one your uptime report does not contain.
First-attempt session success rateBy charger and by connector. The honest reliability measure, and the gap between it and reported uptime is where your missed routes come from.
Departure readiness rateVehicles reaching the state of charge their next route requires. The operational outcome measure — a completed session that ended at sixty percent still failed.
Faults per charger and per manufacturerWhich units and which brands fail most across your sites. The procurement input, and it costs nothing beyond keeping the records.
Charger utilisation by bayHow much of the available dwell each port actually delivers. Reveals whether a queue is a capacity problem or a scheduling one.
Peak site draw against contracted maximumWhere demand charges originate. Visible in the session data long before it appears on a bill.
Energy cost per vehicle and per mileAllocated from session records rather than estimated from a site meter, so it is comparable with fuel cost on the combustion half of the fleet.
Frequently Asked Questions
Our network reports 99% uptime. Isn't that fine?
Uptime and session success measure different things. Industry data shows networks reporting uptime in the high nineties while only around 71% of charging attempts succeed on the first try — because a charger with a frozen screen, a failed authorisation step or a damaged connector is still online and still reporting itself available. For a depot with tight schedules and no margin, the first-attempt success rate is the number that predicts whether routes get covered.
Why does storing failed attempts matter so much?
Because a system that logs only completed sessions makes a repeatedly failing charger invisible. No record means no report, no work order and no pattern — the unit simply keeps failing until somebody complains. It also means you cannot explain a past outage to stakeholders or demonstrate that an operational fix worked. Complete session storage including failures should be contracted explicitly, since reviews of current procurement specifications frequently find it missing.
Which protocol version should we specify?
Open Charge Point Protocol compliance is the non-negotiable baseline for commercial fleet charging. Version 1.6J is the minimum; 2.0.1 is preferred for new deployments because the earlier version lacks the smart charging profile granularity required for departure-based scheduling at scale. Also confirm the platform has been commissioned against your specific charger models — protocol support varies in practice, and a general compatibility claim is not the same as a tested one.
Do we need more chargers, or better scheduling?
Session data is what tells you. Vehicles queuing can mean insufficient ports, or it can mean everyone plugs in simultaneously when staggering sessions would let one charger serve several vehicles across a long overnight dwell. Unexpected demand charges usually indicate the same simultaneity problem. Getting this diagnosis right matters, because grid capacity constraints can turn an infrastructure expansion into a delay measured in many months.
Should chargers be on a maintenance schedule?
Yes — connectors, cables, cooling and enclosures all wear, and repeated connector errors form a pattern well before they cause an outage. Treating a charger as a maintained asset with fault history, scheduled attention and reliability tracked by unit and manufacturer is the difference between a repair list and an emergency. Charging infrastructure downtime carries real industry-wide cost in truck rolls and lost availability; scheduled attention is considerably cheaper.
Can we manage electric and combustion vehicles together?
You should. Most electrifying fleets run both for years, and splitting them across separate systems duplicates every problem — two asset lists, two maintenance queues, two cost pictures that cannot be compared. One dashboard showing battery health on the electric units and fuel consumption on the rest keeps the comparison honest, which is what the next procurement decision depends on.
Where should we start?
Calculate your own first-attempt success rate from a week of session logs, broken down by charger and connector. It takes very little effort and it is almost never the number on the uptime report. In most depots the failures concentrate in a handful of ports rather than spreading evenly, which turns an apparent reliability crisis into a short repair list. Book a demo with a week of session data and we'll run the calculation with you.
Measure the Session, Not the Uptime
First-attempt success calculated per charger, failed attempts stored rather than discarded, error codes raising work orders with history attached, load and scheduling separated from genuine capacity limits — and the whole thing on the same platform as the vehicles it charges.
Works with open-protocol chargers across major manufacturers · Bring a week of session logs to the call · No credit card required
August 25, 2026
By Alex Rowan
All Articles
