Sunday, January 18, 2015

IBM's Node-Red and HA

As I've noted in previous blog entries I've been busy with HA research on my tablet. The research is for my TCF presentation in March and my VCF presentation in April. Actually you could say that I've been a bit hyperactive. I've been busy with work, taking a new course a month, writing several pieces of software, researching topics and preparing the presentations. Much of my research will be posted to my Linux HA and US Home Automation web sites, something I haven't really been doing that well lately.

While researching MQTT I came across some references to Rules Engines to make decisions based on conditions (typical user's if-then-else type stuff used in a lot of HA). One of the things I've come across is Node-Red. A project that was created at IBM for graphically programming the IoT (HA is a subset of the IoTs). At first I thought it was a bit hokey, I recall Logo and a few other similar UIs. This time I think someone got things a lot more right. I was playing with Node-Red and was able to get data from one MQTT topic, parse it in Javascript, and send it back to another MQTT topic. I spent more time trying to recall the Javascript syntax than haggling with Node-Red to get it to work. While I'm not sure I'd try to write a full Linux file server with Node-Red I can see using it for handling the typical HA type Business Intelligence. I would expect the typical home user BI to be less complex than what I see at work but there are similarities. At this moment I'm not sure if it is possible to build the building blocks that would make HA more flexible for the home user but I am sure that by using MQTT, with DHCP, Node-Red and a HA server package it should be possible to have a user add a new node to the existing HA system with a minimum of trouble.


Post a Comment

<< Home