Machine Learning for Manufacturing
Machine Learning for Manufacturing (MFG ML) is a end-to-end
solution accelerator that builds and
deploys a machine learning model for tool wear detection using
Manufacturing Data Engine
and Vertex AI.
Architecture
The Machine Learning for Manufacturing solution accelerator
includes the following components:
-
Manufacturing Connect and Manufacturing Connect edge
- Factory edge platform co-developed with Litmus Automation
that connects to manufacturing assets and streams the asset telemetries to
Manufacturing Data Engine.
-
Manufacturing Data Engine
-
Pub/Sub: message broker for
factory telemetries and integration point for factory edge platform and MDE.
-
Dataflow: fully managed and unified
stream and batch data processing pipeline.
-
Cloud Storage: object storage for
structured and unstructured factory data.
-
BigQuery: serverless data
warehouse for storing structured and semi-structured factory data.
-
Cloud Bigtable: fully managed NoSQL
database optimized for storing factory time series data.
-
Vertex AI
- Unified AI platform for building and deploying machine learning models.
-
Looker
- Business intelligence platform optimized for Google Cloud integration.
Deployment
For instructions to deploy this solution accelerator, please see
deployment.md.