The problem with two data sources

For home-charging reimbursement, Pipelet today reads charging sessions from two places: the wallbox itself (via OCPP 1.6/2.0.1) and the vehicle (via OEM CarData APIs like BMW, Tesla, Volvo, Audi, Ford). For a large fraction of real-world drivers this is enough.

But it misses one growing segment: customers whose wallbox is bound into a home energy management system via the solar-inverter manufacturer's hub — Fronius Wattpilot, Kostal Enector, SMA eCharger, Huawei Smart Charger, EcoFlow PowerPulse. These wallboxes often don't speak OCPP to the outside world; their session data lives inside the inverter cloud. Without an integration, those drivers can't get their home-charging kWh reimbursed at all.

The DACH market in one table

Eight manufacturers cover the vast majority of installed residential PV in Germany and Austria. Five of them ship a branded wallbox that integrates only with their own inverter ecosystem.

Manufacturer Global share 2026 DACH presence API access Own wallbox?
Huawei (FusionSolar) 29 % Very high OAuth2, Token-Request ✓ Smart Charger
Sungrow (iSolarCloud) 23 % High App-key request ✓ iEV
Ginlong/Solis 8 % Medium API-Key Announced
Growatt (ShinePhone) 6 % High API-Key Since 2025
GoodWe (SEMS) 5 % High License + API key ✓ HCA
SMA < 5 % Very high (DE-native) Modbus + Sunny Portal ✓ eCharger
Fronius < 5 % Very high Local JSON + Solar.web ✓ Wattpilot (multi-vendor)
Kostal (Plenticore) < 5 % High Local JSON ✓ Enector (PV-only)
SolarEdge ~10 % High Public Cloud API + Modbus ✓ EV Charger
Enphase small, growing Growing Public Enlighten OAuth ✓ IQ EV Charger
Solarwatt small DE-focused Local REST on Manager
EcoFlow (PowerOcean) small, growing Growing Access Key / Secret ✓ PowerPulse 2

Sources: published Q1-2026 market-share reports (Wood Mackenzie, IHS Markit) and manufacturer documentation. Numbers are inverter-shipment-weighted, not installed-capacity-weighted.

Three camps, three integration strategies

Once you look at the API access patterns, the market falls into three clearly distinguishable camps. Knowing which camp a manufacturer belongs to tells you up front how much engineering effort an integration will cost and how reliable it will be in production.

Camp 1

Walled garden

Huawei · Sungrow · RCT Power · E3/DC · EcoFlow

Closed ecosystem. Best out-of-the-box integration with their own wallbox, worst third-party device compatibility. Cloud-only access by design — local Modbus exists but is unofficial.

Pipelet take: Cloud-API integration only. We pull the inverter's view of its bundled wallbox; the customer cannot mix in a different OCPP wallbox without losing the integration.

Camp 2

Open systems

Fronius · Kostal · GoodWe · FoxESS · SolarEdge · SMA · Solarwatt

Multi-vendor by design. The inverter speaks Modbus or a documented JSON interface locally; the wallbox can be a third-party OCPP device. Cloud APIs exist but local is the gold standard.

Pipelet take: Dual integration: local for privacy-conscious customers (Modbus or local REST), cloud as fallback. The customer keeps freedom to swap wallboxes; we read the inverter purely for surplus-PV context.

Camp 3

Aggregator-served

All of the above via Enode · Sunvoy · evcc

A single API surface in front of 20–30 manufacturer clouds. Avoids per-brand engineering at the cost of vendor lock-in and per-device OPEX. Best for tail brands.

Pipelet take: Tail-cover only. We build Camp 1 + Camp 2 ourselves (top 80% of DACH market share); for the long tail of niche manufacturers we lean on Enode or Sunvoy so we don't tie up engineering on single-digit-percent share brands.

Why this matters for Pipelet customers

A fleet manager rolling out home-charging reimbursement to 500 drivers needs the reimbursement flow to work for every driver, not just the ones with an OCPP-capable wallbox. Today, somewhere between 15 % and 30 % of new German residential PV installs come bundled with a brand-locked wallbox that does NOT speak OCPP to the outside. Without inverter-cloud integration, those drivers get stuck.

For the fleet operator the integration also unlocks a second value: PV-surplus charging as a feature. With knowledge of live inverter production, the wallbox (whether OCPP-controlled or inverter-bound) can be told to draw only when the household has solar surplus — meaningful in the post-EEG-feed-in-tariff era where surplus PV is sold cheaply but consumed at retail rate.

How Pipelet turns inverter data into charging sessions

The integration is more than a raw data pull. Pipelet runs a brand-agnostic reconstructor that walks the periodic inverter-cloud snapshots and turns them into discrete charging sessions in the same shape OCPP wallboxes and OEM CarData produce — same database table, same fleet-API surface, same reimbursement engine. From the operator's perspective, a kWh that came from an inverter-bundled wallbox looks exactly like one from an OCPP wallbox.

┌──────────────────────┐    ┌─────────────────────┐    ┌─────────────────┐
│  Inverter cloud      │    │ Pipelet driver-     │    │ Fleet API consumers│
│  Enphase Enlighten,  │ ─> │  portal worker      │ ─> │ (CF4, FleetX,   │
│  Huawei FusionSolar  │    │                     │    │  partner BIs)   │
└──────────────────────┘    │  c4_pv_snapshots    │    └─────────────────┘
                            │  + reconstructor    │
                            │  ↓                  │
                            │  op_charging_       │
                            │  sessions           │
                            │  id_tag='pv-…'      │
                            └─────────────────────┘

Charging sessions reconstructed from inverter data carry the marker id_tag = "pv-<brand>" in our schema — for example pv-enphase or pv-huawei. Partner portals, dashboards, and the driver portal render those rows with a green "Solar · <Brand>" badge so users see the provenance at a glance. Developer-portal docs describe the markers, the reconstruction algorithm, and how to add a new brand yourself in three files.

Pipelet's roadmap

We're rolling integrations out market-share-first. The phases below are concrete project gates, not aspirational marketing.

Phase What DACH coverage Status
0 Solarwatt + Enphase + PV-Polling ≈ 5 % live
0+ Snapshot → session reconstructor (brand-agnostic state machine; Enphase IQ EVSE wired) foundation for 1–4 live
1 Huawei FusionSolar + Sungrow iSolarCloud + 52 % global market share in clarification
2 Solax + GoodWe + 10 % planning
3 Fronius + Kostal (local) DACH power-users planning
4 EcoFlow PowerOcean + Tail-Cover via Enode/Sunvoy long tail evaluation

If you operate a fleet whose drivers are spread across multiple PV brands and you want a single home-charging-reimbursement flow that just works: talk to us →