LUMINAI Refinery Advisor on AWS
Publication date: February 29, 2024 (Diagram history)
This reference architecture illustrates a highly scalable deployment of Beyond Limits’
LUMINAI Refinery Advisor
LUMINAI is an AI-driven decision support system that maintains a holistic view of the entire site and makes timely, transparent recommendations that enable operators to close the profitability gap in real time and maximize operational efficiency.
LUMINAI Refinery Advisor on AWS diagram
-
Operations data from multiple refineries stream into the LUMINAI Refinery Advisor System on AWS.
-
Application Load Balancer on a public subnet directs data stream to the sensor processing pipeline on a private subnet.
-
Stream data ingestion processes running on Amazon ECS direct streams to Amazon Managed Streaming for Apache Kafka (Amazon MSK), which ingests and processes streaming data in real time at scale and directs it to ETL jobs running on Amazon EMR. As each job completes its task, results are stored on Amazon Aurora PostgreSQL and placed on Amazon MQ, a highly performant and scalable message queuing service.
-
LUMINAI’s Orchestrator and Reasoner business services run on Amazon ECS and process messages on Amazon MQ. Both business services leverage AI-driven algorithms to provide real-time insights into each refinery.
-
Processed data from business services are stored on Aurora PostgreSQL and are made available to web UI, API, and BI tools. Long-term data is stored on Amazon Simple Storage Service (Amazon S3).
-
Users can access the refineries' operational data and have a holistic view of each refinery through a web browser. LUMINAI provides access to refinery data through a web UI or APIs.
-
Authorized clients such as plant and operations managers, maintenance teams, and data scientists and engineers can access information by using BI tools.
Further reading
For additional information, refer to
