Skip to content

Engineering

Example cases

Typical projects, anonymous until customers allow their names, structured as challenge, existing systems, EdgeTech integration, technical solution and result. EdgeTech supplies the technology, control logic and integration that makes chargers, inverters, batteries, meters and heat pumps work together, not necessarily all the components.

View as MarkdownUpdated 2026-09-29

Typical projects

Each will be written as Challenge → Existing systems → EdgeTech integration → Technical solution → Result. Drafts are in progress for:

  1. A European EV charger OEM looking for an open 22 kW charging platform.
  2. A residential energy project combining EV charging, solar and dynamic load balancing.
  3. A smart-home integration combining charger, smart meter and heat pump.
  4. A multi-residential / apartment charging and energy-management project.
  5. Temporary charging and local energy management for a construction site.
  6. A European OEM requiring a customized white-label charger.
  7. A HEMS project integrating EV charging, solar, battery and heat pump.
  8. An OEM requiring custom firmware, Modbus, Matter or API integration.
  9. An energy-management project where existing third-party components must work together.

Five engineering cases, already written

Waking a sleeping EV, on an 8-bit microcontroller

Constraint. A 2022 premium EV stopped resuming charging after a pause of more than a few minutes: it went into deep sleep and ignored the pilot signal. The charge controller runs on an 8-bit microcontroller with most of its flash in use, and the fix could not change behaviour for the hundreds of other vehicle models already charging fine.

What we did. Implemented the control-pilot disconnect wake-up of IEC 61851-23 Edition 2, made it configurable per socket because two other brands reacted badly to it, and rolled it out over the air to the whole fleet. Shipped as firmware 2.10.17 in April 2025.

Moving a live fleet off a vendor cloud

Constraint. Hundreds of site controllers in the field depended on a third-party IoT cloud for connectivity, OTA and events. The migration had to happen without a site visit, without a firmware version that could brick a device, and with the vendor’s cloud switched off per device only after the new path was proven.

What we did. Built an MQTT broker with a per-device key derived from the device ID, a forwarder that turned device events into the webhooks the platform already consumed, and a staged rollout with a per-device ledger and an hourly watchdog that enforces the intended state. The fleet moved in the summer of 2026 with no truck rolls.

Load management without a data cable

Constraint. Charge points on holiday parks, construction sites and large car parks need a grid limit, but running RS485 to each of them was impossible or too expensive. The solution had to work without network infrastructure and fail safe when the radio link was lost.

What we did. With ElaadNL, built the GridShield controller and receiver over LoRa point-to-point, with Modbus, relay and SG-Ready outputs and a configured safe limit on radio loss, and tested it against ElaadNL’s implementation in both directions.

A charging platform a telecom operator can host

Constraint. Corporate customers wanted the charging platform inside their own network, passing their own vulnerability scans, with no dependency on the vendor’s cloud.

What we did. Ported the platform to a Quarkus application with file-based databases, daily backups, TLS from PEM files, a signed licence with a grace period, and a mobile API compatible with the existing apps. It runs on a single Linux host.

A Matter charger before ecosystems support it

Constraint. Home energy management is moving to Matter, but the Matter Energy EVSE device type was new and no ecosystem could commission it. We wanted a charger ready the day one could, without changing the charge controller.

What we did. Built a Thread module that reads the charge controller over Modbus and exposes the Energy EVSE, Energy EVSE Mode and Electrical Power Measurement clusters, commissioned it into our own controllers, and drive it from an agent interface that sets the current limit within the load-balancing envelope.