Dr. June Andrews, Principal Data Scientist, Wise.io, From GE Digital
summary
Claims a deployed real-time production anomaly detector that ingests aircraft engine telemetry, flags out-of-normal conditions, and surfaces a human-attributable feature string pointing to the most likely sensor. Models learn from monitoring-team confirmations and disagreements for continuous improvement. Constraints include limited costly data so feature engineering compares paired engines and tunes metrics for the long tail. Notes that GE covers 60% of airplane engines and 1/3 of world power, and it takes about 5 months to deploy to aviation with transferable gains for power plants.
AI-written summary · about
chapters (27)
- 0:00Introduction
- 0:41What is intuitive machine learning
- 1:05The Boehm Spiral
- 1:59Planning
- 2:38Building
- 3:07Solving Problems
- 4:09evangelize
- 4:27release
- 4:54recurrent process
- 5:39success metrics
- 6:02safety
- 6:12success metric
- 7:37reusable
- 8:51industrial IoT
- 9:34G9115V
- 9:51Fan blades
- 10:55Sensors
- 12:38Realtime in production
- 14:05Compliance to regulations
- 15:12Safety reliability efficiency
- 15:50Big data
- 16:08Be clever
- 16:24Feature a stage
- 17:20Google brain
- 17:46Bob
- 18:01Bob Test
- 18:34Industrial Internet of Things
related talks
30mPredictive Maintenance and Anomaly Detection for Wind Energy
29mHow Does Data Science Revolutionize the World of Machines?
20mHank Roark, Data Scientist, H2O
22mEngineering Predictive Platforms
16mEnterprise and industrial AI: The great threat of complacency
21mDataEngBytes 2023 - SYD-T2-04 - Suneeta Mall
29mAI-Powered Anomaly Detection in Cloud Data Pipelines
23mDataEngBytes 2023 - MEL-T3-01 - Rajath Vanikul