This sample application includes a Node-RED dashboard that can run on a laptop (x86 Mac, Linux) integrated with an IBM i back-end.
- IoT & Event Processing & dashboard with micro-services running on IBM i : ActiveMQ (MQTT broker), Kafka...
- Db2 for i : Nosql/JSON processing & Geospatial Analytics (IoT JSON data in DB2 + real time distance computation/geofencing)
- RPG : Core Business Rules & integration with Kafka (to do - see this IBM Developer article for more inspiration)
- Video trailer & short demo here on BOX
-
Install & start Node-red, import the project (project mode) or the flow
-
Install the flow dependencies and configure the connections to IBM i : mqtt nodes, and ODBC nodes with proper IP address (ODBC config with proper DSN in .odbc.ini file to be done on the machine running the node-red dahsboard). You'll need a Webcam on your laptop and the Tensorflow node installed
-
On IBM i : Install the DB2 for i part from the DDL file provided. Use IBM ACS to run these SQL queries that will create the schema and necessary tables. Install a MQTT broker somewhere - Here , ActiveMQ on IBM i (7.4+ for geospatial analytics)
yum install activemq && /QOpenSys/pkgs/bin/activemq start
- Optional: Raspberry PI Camera: import the Rapsberry PI flow , once the integrated camera plugged and activated. Start Node-red, control that the flow can communicate with the MQTT Broker. Tested on Node.js 10 on PI (tensorflow.js constraint). If necessary , downgrade your Node.js version usign n .
sudo npm install -g n
sudo n 10.16.0
- Go to the Node-RED Dashboard page on your laptop and see the frames and AI inference results streaming to your IBM i.
Author: Benoit Marolleau benoit.marolleau@fr.ibm.com , please reach out for any questions or a feedback!!