In just 45 minutes, John walks you through an end-to-end cognitive IoT solution that shows you how you can take business critical actions on real-time sensor data and then find insights in stored data sets.
In this video:
- John Walicki, Watson IoT Developer Advocate, IBM
John takes you on a fast-paced developer journey that helps you gather IoT sensor data, perform analytics on the edge, gain real-time insights on the data in the cloud, and perform data analytics and visualization on historical data. With IBM Data Science Experience, you don’t need to be an expert data scientist to discover insights in your IoT sensor data using Watson IoT Platform.
Phase 1 – Edge analytics
After connecting the Arduino to the sensors and to the Intel NUC Gateway, John shows the IoT Gateway Developer Hub, looking at the device and sensor data at the gateway.
Node-RED is an open source visual programming tool that comes installed on the Intel NUC Gateway. John shows the Node-RED flows that gathers the various sensor data, and sends it to the Watson IoT Platform quick start.
Phase 2 – Real-time insights in the cloud
John shows how to interact with the data in the Watson IoT Platform Quickstart.
Then, John introduces the IBM Bluemix cloud platform, and the idea of boilerplates being patterns to use to set up a typical IoT app using the Internet of Things Platform Starter boilerplate.
John shows how you can use Node-RED in your Bluemix IoT app to build a flow that filters some of the sensor data and stores it in a Cloudant database. He then shows off the Watson IoT Platform dashboard, and the various boards that help you analyze devices, their data, and your usage of the Bluemix services.
Phase 3 – Saving data in the cloud
Still working in the IBM Bluemix cloud platform, John shows the IoT sensor data collected from the Intel gateway being stored in the Cloudant NoSQL database.
Phase 4 – Data analytics and visualizations in the cloud
Finally, John dives into even more data analytics and visualization by exploring with IBM Data Science Experience, loading the historical data into a Jupyter notebook and using the open source PixieDust tool to visualize the data to gain insights into the IoT sensor data.
Still want to learn more?
John recommended working through Romeo Kienzler’s developerWorks tutorial, Build a cognitive IoT app in just 7 steps, which steps you through many of the phases presented above.
And, if you still want more, John recommends taking the Coursera course that Romeo developed, A developer’s guide to exploring and visualizing IoT data.