The architectural thesis advanced in the preceding chapter resolves, in implementation, into a single deployable object and the system that learns from it. This chapter sets out the iONE node as built — its envelope, components, and field economics — then the layered architecture that holds its commodity inputs interchangeable and its independence structural, the iONEOS control system that makes the node predictive rather than merely autonomous, and the fleet across which that intelligence compounds into a proprietary, durable asset.
1. The Deployable Node
The physical layer is the structurally rigid, thermally and mechanically defined hardware envelope of the iONE node — the aerospace-grade aluminium frame and tracker structure, the IP65-rated cabinet hosting the battery core, the asymmetric single-side thermal architecture, the dual-axis tracking kinematics with three-tier storm protection, and the foundation-free deployment configuration through which the unit reaches operational readiness in three to four hours under a two-person crew. The physical layer is engineered for a twenty-five-year structural design life across the operational climatic envelope of European, Mediterranean, and Gulf deployment zones, on which the foundational fifty-year hot-dip galvanised screw-pile deployment is constructed.
Within that envelope — the slot of the Chapter III pivot — the platform integrates standardised, best-in-class components as the building blocks of every configuration. The envelope sets the permanent form factor, the structural mass, the thermodynamic boundaries, and the operational profile of the station. The components — photovoltaic modules, battery cells, power electronics, and battery management platforms — are treated as interchangeable inputs that comply with the baseline physical interfaces of the envelope. The architecture is the durable, long-term element of the design; the components are the variable.
This structural separation operates across each component class of the iONE node, creating an institutional-grade asset framework.
The Product Family
The node is configured, not bespoke. A single architectural envelope spans two deployment lines — Home and Industrial — across three capacity points (iONE Core 16, 32, and 48 kWh, one to three 16 kWh LiFePO4 modules), with the 720 W tracked solar array scaling from four to six panels. Site parameters are entered into the live configurator at gtlab.org, which resolves them against PVGIS irradiance data (European Commission JRC) for the deployment latitude and returns a costed, generation-modelled station specification in real time — the same customer-facing data-acquisition surface documented in §3. Full renders, dimensions, and per-variant specifications are detailed in the iONE Continental Product Manual (V2.0) and the GT Battery Specification appended to this memorandum.
Photovoltaic Modules
In the current industrial specification, the array utilises 720W Bifacial TOPCon glass-glass panels featuring an Atomic Layer Deposition (ALD) Al2O3 aerospace coating for sand, salt, and UV resistance. Sourced from German manufacturing partners — Heckert Solar or Sonnenstromfabrik — the current modules deliver a cell efficiency profile above 22% and a 30-year engineered glass-glass life free from backsheet degradation. The envelope is natively forward-compatible with the European perovskite-on-silicon tandem pipeline — commercial modules already shipping in 2024, higher tiers maturing through 2032 — Oxford PV's Brandenburg commercial line and the Fraunhofer-certified 26.9% module already among its outputs — absorbing them without structural revision; the technology trajectory is documented in Chapter V.
Battery Cells and Electrochemical Core
The battery core operates on prismatic Lithium Iron Phosphate (LiFePO4) chemistry utilising the industry-standard 314 Ah industrial format. The cell substrate is dimensionally and electrically interchangeable: while the current validation unit is architected around XDLE (Xingdong Lithium Battery) CBA71173204-314Ah cells, the internal envelope is fully compatible with cells from the Hungarian gigafactory cluster, emerging MENA-based manufacturing initiatives, or tier-one alternatives (CATL/EVE-grade) without structural modification. The platform remains supplier-agnostic; the architecture survives any one geopolitical configuration of the cell supply chain.
Power Electronics and Redundancy
MPPT solar chargers, inverter modules, and supercapacitor buffers comply with industry-standard electrical and communication interfaces, deployed with N+1 hot-swappable operational redundancy across both the 48V DC telecom-grade bus and the AC inversion paths. The architecture supports dual supply tracks: cost-optimised global components for civil applications, and premium European vendor modules (CE+T Power, Eltek, Benning, Vertiv) for resilience-critical, assured-supply infrastructure.
Structural and Thermodynamic Foundation
The physical envelope is constructed from aerospace-grade 6061-T6 aluminium for the main frame and tracker, marine-grade Stainless Steel 316L for fasteners and chloride-exposed joints, and hot-dip galvanised steel for the foundation — guaranteeing a 50+ year ground life and a 25-year structural design life across aggressive environments. The iONE Core Battery Module is housed in a standalone, IP65-rated outdoor cabinet (300 × 250 × 1280 mm) under a disciplined, asymmetric, single-side active thermal architecture: active heating and phase-change buffering occupy one side of the cell pack opposite the system electronics, with heat distributed uniformly through the stack via high-conductivity copper busbars between a silica-aerogel insulation layer and a monolithic Foamglas cellular-glass block on the opposite face. The full thermal-stack specification — phase-change material, insulation, sealing, and the dual-stage pressure-vent and outgassing system — is detailed in the iONE Continental Product Manual (V2.0) appended to this memorandum.
Kinematics, Protection, and Deployment Economics
The precision dual-axis tracking system operates at ±0.1° via self-locking worm-gear slewing drives with 4-quadrant photodiode and GPS-ephemeris guidance, under a three-tier predictive storm-protection framework — continuous tracking below 60 km/h, a vertical segmented drop rated to 180 km/h, and a full book-fold survival mode for sand, snow, and extreme high-altitude or coastal conditions — detailed in the Product Manual.
Every unit is installed via a foundation-free anchor configuration comprising four 1.0-metre perimeter frame anchoring piles and two 2.0-metre centre mast piles (76 mm shaft, 200 mm helix) driven by a handheld hydraulic head. This eliminates concrete works, ensuring complete assembly, commissioning, and cloud registration within 3 to 4 hours by a two-person crew, while enabling swift site relocation with zero permanent land modification.
Two operational prototype stations are currently undergoing active field validation at the Berlin test site. Complete mechanical tolerances, single-fault tolerance mapping (IEC 62619 §8.3), and component-level specifications are detailed in the iONE Continental Product Manual (V2.0, January 2026) and GT Battery Specification GT-BAT-SPEC-001 (V1.3, May 2026) appended to this memorandum.
Transparent Compliance Pipeline
The platform is engineered for full alignment with the EU Battery Regulation 2023/1542 (carbon footprint declaration, digital battery passport, and upstream due diligence frameworks). The compliance pipeline — third-party LCA verification through specialised consultancies (Sphera, Quantis), digital passport implementation via web-accessible QR codes on the laser-engraved FEM marking plates, and component-level IEC 62619, UN 38.3, and RoHS safety certifications for the 314 Ah cell variant — is scheduled for complete formal execution before the first commercial shipment series leaves the Berlin facility.
2. Architecture, Not Atoms
The platform's architectural integrity rests on the operational separation of three layers whose engineering, commercial, and contractual properties are distinct: the physical layer, the internal control layer, and the external orchestration layer. The separation is engineered into the platform from the design phase forward, not asserted post-construction as an architectural rationalisation. The consequences across the platform's commercial trajectory are direct.
The internal control layer is iONEOS, the deterministic-and-probabilistic intelligence system that governs the operation of the physical layer. The internal layer commands tracker kinematics, manages cell-level battery state across the prismatic LFP cell array of the iONE Core configuration, regulates the thermal envelope across the silica-aerogel and phase-change buffer, processes fault signals across the redundant power-electronics paths, and structures the telemetry stream that the node generates across its operational life. The internal control layer operates autonomously across the platform's design envelope; it does not require external connection for its operation; it does not depend on any single orchestration platform or any single commercial counterparty for its functional integrity.
The external orchestration layer is the protocol-and-standards surface through which the deployed node integrates, at the operator's election, into the European distributed-energy market — EEBus, Modbus TCP, OCPP, SG Ready, and the parallel set of European industry-standard interfaces on which the orchestration platforms operating across the European market are constructed. The external layer is the layer of optional market participation, never of operational dependency. The deployed node functions identically in the off-grid critical-infrastructure configuration where the external layer is inactive, and in the grid-connected market-participation configuration where the external layer is active through the operator's elected orchestration platform.
The commercial consequences of the separation are direct. The physical layer is the unit-economics base of the platform: the hardware margin captured at the point of sale. The internal control layer is the recurring-revenue base: the iONEOS subscription captured across the operational life of the deployed unit. The external orchestration layer is the grid-flexibility revenue base: the dynamic-tariff and §14a EnWG participation captured at the grid-connected fraction of the fleet. The three layers correspond directly to the three revenue layers of the economic architecture documented in Chapter VI. The architecture is not a technical preference; it is the engineering precondition for the commercial thesis.
Open Seams: Compatible by Standard, Independent by Design
The boundaries between the three layers of the iONE architecture — physical, internal control, external orchestration — are governed by standardised European protocols rather than proprietary integration code: EEBus, Modbus TCP, OCPP, and SG Ready define the communication surface at every functional seam. There is no proprietary integration debt anywhere in the architecture. Independence by design, compatibility by standard.
This produces three structural properties. First, deployment independence: off-grid, the node operates as a complete standalone asset with no external dependency; grid-connected, it enrols into market participation through any orchestrator that implements the standards — gridX, Octopus Kraken, 1KOMMA5°, Tibber, Next Kraftwerke — without bespoke integration, and can be removed without functional consequence. Second, component-level supplier substitution: Hungarian cell capacity, European tandem modules, or alternative power-electronics vendors are absorbed without revision of the control or orchestration layer. Third, operational independence from any single technology partner: the failure or strategic redirection of a platform does not constitute risk to the installed base.
The position is durable across twenty-five years of operational life: component generations and orchestration platforms will turn over multiple times, while the architecture — defined by its envelope and its open communication surfaces — persists without structural revision. The principle is the inverse of integrated-product strategy: Apple iPhone integrates components into a vertically engineered product; iONE integrates an envelope into which components are inserted across generations. For infrastructure with a twenty-five-year design life, open seams are the only architectural choice that survives the supply chain its components will pass through.
3. The Infrastructure Brain: iONEOS
iONEOS operates today across a defined functional architecture. The customer-facing deployment-planning layer is live on the public configurator at gtlab.org — resolving customer-specified parameters (location, product line, energy requirement, component selection) into a defined station configuration with transparent pricing and financing, and operating as a structured data-acquisition layer addressable through customer interaction today.
The on-station control logic is structured into two operational regimes. The deterministic core handles battery state-of-charge boundaries, cell-level voltage and temperature monitoring, thermal management commands (heater activation below 0°C charging threshold, PCM phase-change buffering across the operational range), MPPT control across the redundant solar charging paths, tracking command sequences against astronomical ephemeris and photodiode feedback, storm protection triggers against ultrasonic wind data and weather-API forecasts, and fault containment routines isolated to the affected subsystem. This layer is designed for hard-coded, auditable execution against the engineering envelope of the node; its behaviour does not depend on machine-learning inference and remains deterministic across the operational life of the unit.
The probabilistic intelligence layer activates as fleet-scale telemetry accumulates beyond statistical thresholds required for model validation. Predictive maintenance signatures, degradation curves for the specific cell class deployed in a specific climate zone, optimised tracking adjustments under recurring local irradiance patterns, market-arbitrage signal processing for grid-connected nodes — all of these are model-based outputs that enter operational control as the fleet data supports them. The probabilistic layer is developed in scientific consultation with Vladislav Andrushko, doctoral candidate at the Ubiquitous Knowledge Processing Lab of the Technical University of Darmstadt (under Prof. Dr. Iryna Gurevych); Andrushko's doctoral research on self-supervised learning for autonomous cyber-physical systems is directly applicable to fleet-scale anomaly detection across the distributed iONE deployment base. The architectural separation is explicit and audited: deterministic logic for safety and compliance, probabilistic models for optimisation; the latter never overrides the former.
The integration of these two regimes follows the standard architectural pattern for autonomous infrastructure systems. Safety-critical and regulatory-binding decisions execute on deterministic logic with full audit traceability; performance-optimising decisions execute on probabilistic models trained on fleet-scale operational data. The boundary is engineered, not negotiated.
The Predictive-Maintenance Loop
What this architecture is built to produce is already demonstrable as a working design. iONEOS resolves the telemetry stream of each node — temperature, wind, state-of-charge, link quality, tracker geometry — into a continuous fleet picture, and is engineered to act on deviation before it becomes failure: a recurring angle mismatch read against bracket micro-shift is classified as tilt-sensor drift, the affected node flagged, the owner notified, and a technician dispatched with the diagnosed cause and the specific part — the predictive-maintenance loop designed to close without human monitoring. A walkthrough of this fleet-operations layer is published at gtlab.org.
4. The Fleet Is the Asset
Each of the preceding sections describes a unit; the asset is the fleet. A single iONE node is a complete, deployable energy station — but its durable value is not the hardware margin captured at the point of sale. It is what the deployed base accrues over time: the probabilistic intelligence layer of iONEOS activates only once fleet-scale telemetry crosses the statistical threshold its models require, and the degradation signatures it learns — specific to a cell class in a specific climate zone — are objects that exist only once a fleet has generated them. The deterministic node ships from unit one; the self-learning asset is built, climate by climate, across the deployed base.
That accrual is the platform's durable moat. The data-sovereignty position established in Chapter III is realised through the probabilistic intelligence layer set out above; what converts the position into a proprietary asset is the multi-climate LFP-degradation dataset no competitor — vendor, orchestrator, or integrator — holds, because none operates an instrumented physical fleet across the European, Mediterranean, Gulf, and Arctic envelopes at once. The event that crosses the activation threshold, and the validation fleet that produces the first such dataset, is the Seed programme set out in Chapter IX. The node is the body; the fleet is the asset; the point at which the architecture begins to compound is the point at which the Seed capital is deployed.
The compounding is platform-shaped, not product-shaped. Every node deployed deepens the dataset and sharpens the next node's intelligence — a flywheel in which the advantage widens with the installed base rather than eroding to competition: the moat is the fleet itself, and no entrant can assemble it retroactively. The same architecture shifts the margin from the hardware to the platform — iONEOS runs for the operational life of every node, GT-built or licensee-built at parity, so recurring value accrues across the entire installed base on the Android pattern, with the originator holding the protocol and the standard while the category scales through others. The first units sell hardware; once the category is the default rather than the exception, it is monetised at the platform layer the hardware was always built to carry.
Bridge to Chapter V
The architecture is documented — the deployable node, its three-layer separation, the iONEOS intelligence that governs it, and the fleet that compounds them into a self-learning asset. Chapter V measures the climate performance of that fleet against the 100 Mt CPP threshold; Chapter VI follows with the economic engine built on the same foundation.