LitmusLitmus Logo

Google Manufacturing Data Engine Reference Architecture

Explore the architecture patterns manufacturers use to deliver contextualized industrial data from edge into Google Cloud Platform and the Manufacturing Data Engine.

Google Manufacturing Data Engine pdf

This architecture shows how Google Manufacturing Connect Edge (MCe), powered by Litmus Edge, connects industrial systems and delivers contextualized data to Google Cloud Platform and Manufacturing Data Engine.

Architecture Summary

This architecture shows how industrial OT data is collected, normalized, contextualized, and delivered from the edge into Google Cloud Platform using Google Manufacturing Connect Edge (Powered by Litmus) and Google Manufacturing Connect.

  • Manufacturing Edge Connect (MCe) connects to industrial systems, normalizes and contextualizes data, and prepares it for consumption — deployable locally or in Google Distributed Cloud (GDC).

  • The Edge Data Foundation with Google Manufacturing Connect Edge (MCe) bridges contextualized industrial data into the Manufacturing Data Engine, enabling integration with Data Engine Pipelines, Storage, and Services.

  • Google Manufacturing Connect (MC) enables centralized deployment, monitoring, and lifecycle management of MCe instances across the factory floor.

  • Google Cloud Platform provides a consistent management plane across edge and cloud for governance, data services, analytics, and AI-driven consumption.

End-to-end data flow

1. Connect and collect industrial data
MCe connects to PLCs, controllers, industrial robots, CNC machines, IP cameras, databases, file systems, SCADA, MES, and QMS systems to collect raw tags, telemetry, and events using OT Native Connectors, device discovery, and tags & events ingestion.

2. Normalize and prepare data at edge
MCe standardizes tags, units, and timestamps, and applies asset & site context so data is reliable before leaving the edge, through tag normalization, unit/timestamp standardization, and data transformation.

3. Store and buffer at the edge
MCe DataHub stores time-series data, events, and messages locally to support resilience and consistent delivery, using a time-series database, event store, and message buffer.

4. Contextualize data for GCP consumption
MCe Analytics and Data Models organize data into asset/process models, contextual data products, and schema-aligned payloads through schema alignment and delivery preparation.

5. Google Manufacturing Data Engine
The Google Manufacturing Connect Edge (MCe) bridges contextualized industrial data into the Manufacturing Data Engine. Data flows into Data Engine Pipelines (Batch Importer, Data Flow Transformation, Cloud Pub/Sub) and Data Engine Storage (Batch Importer, BigTable, Cloud Storage, Cloud Pub/Sub), governed by Data Engine Services (Configuration Manager API, Metadata API, Data Access API).

6. Deliver to Google Cloud data consumers
Contextualized industrial data flows to Google Cloud consumption endpoints including Looker Accelerators, the Enterprise Data Foundation, and Vertex AI Models, feeding downstream Enterprise Consumer Apps such as ERP, CRM, and PLM.

7. Centralized edge management
Google Manufacturing Connect (MC) powered by Litmus, provides centralized device management, application rollout, central monitoring, and multi-site scaling across the fleet of edge instances.

CTA gradient main section