Skip to content

Edge computing

What Edge Computing Is

Edge computing is a computing approach where data is processed close to where it is generated — on local devices, gateways, or on-site servers — instead of sending everything to a centralized cloud first. The “edge” can be an EV charger, a site controller, an industrial PC in a depot, or a local gateway that aggregates multiple devices.

Why Edge Computing Matters

Edge computing improves performance and resilience in systems that need fast decisions and reliable operation, even when connectivity is limited. It helps by:
– Reducing latency for real-time control (faster response than cloud-only)
– Increasing resilience when internet connectivity is unstable or unavailable
– Lowering bandwidth and cloud costs by filtering and compressing data locally
– Improving privacy by keeping sensitive data on-site when needed
– Enabling local automation without waiting for remote commands

Edge Computing in EV Charging

In EV charging, edge computing is commonly used to keep charging sites operational and optimized under real-world conditions:
Dynamic load management running on a site controller to enforce a site power cap
– Local decision-making for dynamic bay allocation and depot priorities
– Buffering OCPP messages and session data during connectivity outages
– Local driver authentication fallback using cached whitelists or tokens
– On-site coordination with DER such as PV and BESS via an EMS
– Faster fault response routines and automated recovery actions for improved uptime

Typical Edge Components

Edge architectures often include one or more of these elements:
Charger controller logic inside the EVSE (local protections, throttling)
Site gateway that aggregates chargers and meters, and enforces rules
Industrial PC / edge server running optimization and integration software
– Local interfaces to meters, CT clamps, PLCs, building management systems
– Secure connectivity modules and certificate management for device identity

What Edge Computing Enables

Edge computing supports capabilities that are hard to guarantee with cloud-only control:
– Real-time control loops for dynamic load throttling
– Predictable behavior during cloud outages via local fallback modes
– Data preprocessing: anomaly detection, event filtering, compression
– Faster integration with local systems (building loads, energy meters)
– Operational continuity for depots where missed charging can mean missed departures

Best Practices

– Define clear boundaries between edge responsibilities and cloud responsibilities
– Use secure device identity: device authentication and certificate-based connections
– Implement safe defaults if edge logic fails (fail-safe limits, local protection)
– Log decisions locally with timestamps for troubleshooting and audit trails
– Plan remote update strategy carefully with rollback options for edge software
– Monitor edge health as part of overall network diagnostics

Common Pitfalls

– Too much logic at the edge without centralized governance and version control
– Poor update and rollback process causing site-wide instability
– Inconsistent configurations across sites leading to unpredictable outcomes
– Weak cybersecurity on gateways (default passwords, exposed ports)
– Assuming edge replaces cloud analytics and fleet-wide optimization rather than complementing it

Dynamic load management
Depot power management
Device provisioning
Device authentication
Digital twins
Diagnostics
Energy management system (EMS)
Downtime optimization