Anomaly Detection
summary
An edge application pairs Intel DAAL with Eclipse Kura on an Intel Atom gateway to detect anomalies in live sensor streams; Java OSGi bundles implement configurable pipelines that ingest accelerometer FFT data, extract median-frequency features, and run an outlier algorithm that builds a 100-point baseline and raises a notification after a 2-anomaly trigger. Demos show vibration monitoring for fans and a two‑hidden‑layer MNIST classifier with pretrained weights pushed to the edge, both running concurrently on the device and producing live detections.
AI-written summary · about
chapters (7)
- 0:00Introduction to edge analytics
- 1:02Edge analytics framework design
- 4:04Pipeline and configuration
- 7:08Technical stack and DAL details
- 9:57Proof of concept use cases
- 15:56Demonstrations
- 26:48Q&A session
related talks
29mREAL-TIME ANOMALY DETECTION MADE EASY
26mAnomaly Detection System for the CERN Cloud
34mHow to Train Your Dragon and Its Friends: AI on the Edge with Eclipse Kura
39mMastering real-time anomaly detection with open source tools
48mPaths to anomaly detection w/ TIBCO data science, streaming on AWS
44mAnomaly Detection in Telecom with Spark
30mAnomaly Detection with Apache Pinot & ThirdEye
52mDevil in the Haystack