Know about failures
before your
drivers do.
ChargeOS monitors every signal from every charger and tells you — in plain language — which ones are at risk, and why. Built for operators running 5 to 500 chargers.
Simple enough to check in once a day.
ChargeOS is built for parking managers and facilities coordinators — not data scientists. Three things happen automatically, every day.
Connect your chargers
Point any OCPP-compatible charger at ChargeOS. No hardware changes, no proprietary adapters. If it speaks OCPP, it works. Every session, every heartbeat, every fault code is captured automatically.
We watch the signals
Every message your chargers send contains signals — voltage trends, temperature readings, error patterns, session quality. ChargeOS runs continuous analysis across all of them, looking for patterns that precede failures.
You get a prioritized list
Your dashboard shows a ranked list of chargers that need attention — with the reason, the severity, and what to do about it. No log files. No noise. Just the things that matter today.
17 signals. Zero black boxes.
ChargeOS monitors 17 distinct signals across safety, hardware, performance, and connectivity. Every alert tells you exactly which signal triggered it, what the value is, and what to do about it.
Ground fault + overcurrent
Safety-critical signals trigger immediate action. The connector is taken offline automatically and your team is notified — no manual monitoring required.
Thermal fault trending
CUSUM tracks whether thermal events are increasing in frequency. A single event is noise — a trend is a maintenance signal.
Energy delivery ratio
EWMA monitors whether a charger is delivering less energy per session than expected. Gradual degradation that drivers won't report until it's obvious.
Session abort rate
Sessions ending within 60 seconds with no energy delivered. Rising abort rates indicate connector wear or vehicle communication issues.
Heartbeat irregularity
Variance in heartbeat timing reveals firmware instability, network degradation, and impending connectivity failures — before the charger goes offline.
Unexpected reboots
Charger restarts without a preceding reset command. Increasing frequency points to power supply issues or firmware bugs.
One number. Full breakdown.
Every charger gets a health score from 0 to 100. It's the minimum score across all its connectors — a single weak connector brings the whole unit's score down. Click any charger to see exactly which signals are driving the score.
Primary driver: voltage drift (+2.3σ over 4 days)
Secondary: 3× VoltageImbalance in last 48h
Statistical process control, not a black box.
ChargeOS uses CUSUM and EWMA — industrial process control methods used to detect equipment degradation in manufacturing for decades. They work from day one with no historical training data, and every alert is fully explainable.
One avoided truck roll pays for 10–15 months.
Simple per-port pricing. No seat fees, no setup costs, no surprise line items. Annual plans get 2 months free.
Full connectivity, live dashboard, and rules-based alerting. Up to 20 chargers.
Everything in Starter, plus the full predictive analytics stack and unlimited chargers.
White-label deployment, OEM-embedded licensing, and custom SLAs for large operators and hardware manufacturers.
| Feature | Starter | Pro | Enterprise |
|---|---|---|---|
| Core platform | |||
| OCPP 1.6 + 2.0.1 | ✓ | ✓ | ✓ |
| Real-time dashboard | ✓ | ✓ | ✓ |
| Remote commands (all 10) | ✓ | ✓ | ✓ |
| Session history | 90 days | 12 months | Unlimited |
| Charger limit | 20 | Unlimited | Unlimited |
| Alerts and notifications | |||
| Threshold alerts (safety) | ✓ | ✓ | ✓ |
| Slack notifications | ✓ | ✓ | ✓ |
| Email alerts | ✓ | ✓ | ✓ |
| Predictive analytics | |||
| Health scores (0–100) | — | ✓ | ✓ |
| CUSUM drift detection | — | ✓ | ✓ |
| EWMA trend analysis | — | ✓ | ✓ |
| Predictive failure warnings | — | ✓ | ✓ |
| Configuration drift detection | — | ✓ | ✓ |
| Organization and access | |||
| Sites and groups | — | ✓ | ✓ |
| Team roles (RBAC) | 1 user | 3 roles | 3 roles |
| Audit trail | — | ✓ | ✓ |
| Enterprise | |||
| White-label branding | — | — | ✓ |
| On-prem deployment | — | — | ✓ |
| Custom SLA | — | — | ✓ |
| Dedicated onboarding | — | — | ✓ |
Annual plans include 2 months free. · One avoided truck roll ($300–500) covers 10–15 months of Starter at 20 ports.
Questions from facilities teams.
Answers to what operators ask before they start.
Which charger brands are supported?
Any OCPP 1.6 or 2.0.1 compliant charger. ChargeOS is protocol-native — there is no vendor lock-in and no approved hardware list. If your charger speaks OCPP over WebSocket, it works. Common brands include ABB, Schneider Electric, Wallbox, ChargePoint, Enel X, Blink, EVBox, Tritium, and Delta.
How long does setup take?
Register your charger in the dashboard (about 2 minutes), enter the generated credentials on the physical charger, and it connects automatically on the next boot. Most installations are live within 15 minutes of starting.
How does predictive maintenance work?
ChargeOS learns each charger's normal behavior over the first 20 charging sessions. After that, it monitors 17 fault and performance signals per connector using statistical process control — specifically CUSUM (Cumulative Sum) for gradual drift and EWMA (Exponentially Weighted Moving Average) for trend detection. When a signal drifts from baseline, you receive a plain-language alert with a specific recommendation, usually days before a failure would occur.
What does "works day one" mean if there is a learning period?
Safety-critical alerts — ground fault, overcurrent, and tamper detection — are active immediately from the first connection with no learning period required. The statistical drift detection (CUSUM and EWMA) activates after 20 sessions per connector, typically 1–2 weeks of normal usage. During the learning period, the health score shows "--" but your chargers are fully monitored for safety events and threshold violations.
How do I receive alerts?
Three channels: in-dashboard notifications pushed in real time (no page refresh needed), Slack webhook (configurable per workspace in tenant settings), and email alerts (configurable recipient list in tenant settings). Every alert includes the triggering signal, its current value versus baseline, the detection method, and a plain-language recommendation for your team.
What data do you store and for how long?
Starter tier retains 90 days of telemetry. Pro retains 12 months. Enterprise offers unlimited retention. All data is stored in your own infrastructure — ChargeOS is self-hosted via Docker. We do not share your data with third parties, and no telemetry leaves your deployment unless you explicitly configure an integration such as a Slack webhook.
Can I deploy on-premises?
Yes. ChargeOS is fully self-hosted using Docker Compose. Your charger telemetry, user accounts, session records, and audit logs stay in your infrastructure. Enterprise tier includes dedicated onboarding support for on-premises deployments, including TLS configuration and network architecture guidance.
Do I need a dedicated IT team?
No. ChargeOS is designed for facilities teams, not IT departments. The dashboard is self-explanatory, alerts are written in plain language with specific recommended actions, and remote commands (soft reset, unlock connector, get diagnostics) let you resolve common issues without visiting the charger or calling a technician.