Event Streaming Architecture for Industry 4.0
summary
A blueprint for end-to-end Industry 4.0 streaming on Cloudera DataFlow connects edge agents to central analytics: Apache NiFi and MiNiFi collect and prefilter sensor data into local Kafka clusters per plant, then central Kafka feeds Apache Flink which reconciles event time, handles late arrivals and maintains state. A demo pipeline runs real-time aggregations and deploys a logistic regression that reaches ~96% precision predicting defective parts, showing the architecture works at scale. Flink Event Time and watermarks enable cross-plant correlation and CEP.
AI-written summary · about
chapters (17)
- 0:00Intro
- 1:10What is Industry 40
- 1:41Typical Use Cases
- 3:31Platform Components
- 5:32Challenges
- 9:32What is Putty
- 11:17Why Putty
- 13:45Why Link
- 16:57Data Source
- 18:31Workflow
- 19:56Demo
- 27:52Data in Kafka
- 28:19Example
- 29:27Histogram
- 30:04Logistic Regression
- 31:05Regression Results
- 32:24Roadmap
related talks
38m7 reasons to use Apache Flink
40mCreating an open, vendor-independent Industry 4.0 solution with IndustryFusion
20mBuilding realtime efficient queries in an IOT platform
56mTaming IoT Data: Making Sense of Sensors with SQL Streaming
01h 15mIoT Realized - The Connected Car v2
31mFun with Sensors
30mHitachi applying AI in Industry 4.0 digital transformation projects
48mNext-Generation IoT Platform Architecture using MicroProfile + NoSQL + Kafka