Data flow overview
Data Infrastructure Diagram and Components
Data acquisition begins usually with components installed on the edge, on premise, on the local factory network. With LogiX we provide proprietary Logix Gate which requires an Industrial PC which:
can handle many communication protocols in parallel - Modbus, OPC, HTTP, SQL. It listen for value changes and send them to our cloud-based processing.
buffers all messages if connection is not available, and send them as soon as it’s back.
sends messages in batches of configurable size.
Alternatively other edge components can be used as long as they produce messages in compatible format, i.e. LogiX can consume messages originating from Microsoft IoT Edge.
Messages format:
Messages from edge are in standardized JSON messages as specified in https://ilabo.atlassian.net/wiki/spaces/LW/pages/1795424257 (delivered from OPC UA PubSub JSON format as specified in part 14 of the OPC UA specification). As stated earlier, messages can be produced by any component other than Logix Gate (i.e. Microsoft IoT Edge).
Data ingest
On cloud side data are received by platform-managed event ingestion service. In most cases it is Azure Iot Hub, which can scale to receive huge amounts of events per second.
Internal Event Bus
LogiX is an event-driven platform: all back-end services communicate using messages exchanged on Azure Event Hub (implementing Kafka APIs).
Data processing
Each signal (value change transmitted from the Edge) can trigger rule evaluation in LogiX Processing Engine. It is implemented as state-of-the-art complex event processing engine and it is intended to process huge amounts of events in near-realtime speeds. Rules can be configured as simple expressions or complex scripts.
Processing rules, signals configuration along with factory structure and supporting parameters are maintained in dedicates LogiX Serivices (LIB/HUB)
Data storage
All data, both incoming raw signals and facts augmented by Processing engine are saved in dedicated storage services: SQL Database for relational queries, Time Series for fast access to temporal data and File (blob) storage for files (i.e. video files)
Domain Services
LogiX backend components supporting domain and client applications are decomposed into separate microservices dedicated to specialized subdomains (ie. Production Order Management, Asset inventory, Reporting/Notfications, User Administration etc). They are deployed as containers using Azure Kubernetes Cluster.
