Applying AI to IoT Data
summary
Edge device is an STM32MP1 running Linux that samples temp, humidity and IMU data and sends JSON every ~5 seconds via Node-RED + MQTT -> Watson IoT -> Cloudant. Training happens in Watson Studio using Spark in Jupyter (linear/logistic regression) on the stored time-series, yielding coefficients that are sent back as JSON over MQTT. Edge Node-RED scores locally and raises an anomaly alert when coefficients indicate overheating — demo works: finger-on-sensor triggers alert.
AI-written summary · about
chapters (13)
- 0:00Intro
- 5:46Demonstration
- 8:53MQTT
- 11:52IBM Cloud
- 15:23RealTime Analytics
- 19:15Cloudant Database
- 22:12Watson Studio
- 23:48Watson Studio Instance
- 24:19Jupiter Notebook
- 28:00Cloud Portal
- 31:04Testing Data
- 33:18Why is this important
- 47:43Recap
related talks
37mFrom Sensors to Insights at the Edge: Applying AI to IoT Data
45mIoT End to End
32mExtending machine learning to industrial IoT applications at the edge with AWS
49mEnable IoT with Edge Computing and Machine Learning
01h 00mIntelligent Interfaces with Cognitive APIs, BOTs and IOT
27mOpen connected healthcare with Red Hat Edge Portfolio
59mConnecting Devices in the Field to Microsoft Azure
18mManagement of IoT TinyML Devices