Scalability in the Aero-Factory is not merely about adding server capacity; it is about the ability of the Digital Highways to absorb exponential increases in data velocity and volume without degrading system latency or semantic integrity.
The architecture employs a "Cellular" growth model. As new airport bays or de-icing trucks are added, the system scales at the Edge (L3) without increasing the computational burden on the Central Hub. Vortex Proxy Sharding ensures that each physical asset is paired with an independent instance for deterministic processing, while Broker Federation distributes the Pub/Sub load across multiple nodes to prevent congestion.
By separating the Global Stream Collector from the Telemetry Normalizer, the system can scale its processing power based on data type requirements. The Ingestion Bridge on L4 uses a non-blocking architecture for asynchronous normalization, allowing telemetry to be processed in parallel during peak bursts. Meanwhile, the Global Stream Collector (L5) utilizes dedicated storage bandwidth to offload heavy binary streams without impacting the performance of process databases.
The persistence layer is tiered to match the natural "cooling" of data, transitioning from hot real-time packets to cold historical archives through different scaling mechanisms.
| Scaling Dimension | Architectural Response | Benefit |
|---|---|---|
| High Velocity (L3.5) | Local Historian Circular Buffering | Maintains sub-second write speeds for raw data regardless of cloud connectivity. |
| High Volume (L5) | Time-Series Partitioning | Optimizes query speeds in the Central Historian by partitioning data by time and Asset ID. |
| Big Data (L6) | Cloud Lakehouse Elasticity | Leverages decoupled storage and compute to store years of data with near-infinite capacity. |
Unlike rigid relational schemas, the Common Data Model (CDM) within the UDS envelope allows for "schema-on-read" flexibility. New sensor types or business attributes can be added to the CDM payload without requiring a complete redesign of the Digital Highway or the existing database structures. This allows the information system to evolve alongside the physical factory.
In the Aero-Factory architecture, availability is defined as the presence and reachability of valid, contextualized information at the right layer of the ISA-95 stack. This is managed through the Digital Highways (the Vortex Data Fabric), which ensures that data flows seamlessly from emission to long-term persistence.
Information availability is realized through two distinct consumption models: Data in Motion (via Pub/Sub on the Vortex Data Broker for real-time dashboards and analysis) and Data at Rest (via standardized APIs for querying).
Data is born at the Edge and immediately broadcast onto the Digital Highways. The availability of "Data in Motion" is the primary concern for real-time operations.
Availability of "Data at Rest" is distributed to ensure site autonomy and forensic reliability. We distinguish between Cumulative Storage (History) and Stateful Caching (Context).
| Level | Storage Component | Data State | Information Purpose |
|---|---|---|---|
| L3 / L4 / L5 / L6 | ERP DDC (Local Cache) | Stateful / Reflected | Localized cache of ERP/PLM data tailored for Zones' autonomies. |
| L3.5 | Local Historian | Raw / Denormalized | High-resolution "As-Is" records for local forensics and Store-and-Forward buffering. |
| L4 | Asset Mgmt Database | Reference Metadata | The "Source of Truth" for equipment identities used to validate UDS payloads. |
| L5 | Central Historian | Raw / Normalized | Synchronized, clock-drift-corrected process data used for site-wide BI and reporting. |
| L5 | Global Stream Collector | Synchronized Binary | Centralized multimedia archives indexed against telemetry for event reconstruction. |
| L6 | Big Data Lakehouse | Raw & Refined / Normalized | Massive-scale datasets optimized for AI model training and global benchmarking. |