Flotilla with Node-RED and home automation

Here is a blog post on my use of the Flotilla mini-kit, linking it to my home automation system, using Node-RED - https://geekgrandad.wordpress.com/2015/12/06/flotilla/.

5 Likes

I can’t <3 this enough. Damned nice work getting it up and running with Node-RED so quickly. I’d done it before so I knew it was possible, but didn’t expect it to be replicated so quickly.

It’s still on my radar to learn about Node-RED to a degree where I can create custom nodes specifically to deal with Flotilla and make the whole thing a little tidier. It should be reasonably straight-forward.

For advanced users, I think Flotilla’s ability to integrate with other systems will be a big part of the appeal. Plug-n-play dials/sliders/sensors/etc.

I’ve been using Node-RED quite a bit lately, and I have been writing Websocket applications, So together with your previous post on using Node-RED, it wasn’t very hard. I had to look at the Rockpool source to see that you needed to send “ready” to the Websocket, but after that the format of the packets was quite easy to understand. A set of nodes that could be installed by npm would be really nice.

@lawrie May I ask you to kindly share the flow you used in your blog post? either here or on flows.nodered.org it could help jumpstart early node-red integration.

Thank you so much.

I’m not an OP, but I implemented something similar. Here is the flow.

1 Like

Cool. I’ll take a look when I’m near my node-red install. Thank you!

edit: ok I did a quick glance with Pagenodes and I can see the flow is a combination of the previous blog/forum posts.

If I can, my idea is to wrap up some subflows (each for one of the modules) with the splitted outputs.

Thinking about developing the “real” node-red flotilla nodes, I’ll vote to use the websocket approach (instead of using the direct usb-serial protocol and sorta rewriting rockpool in node-red)
Doing so, you can decouple the host where node-red is running and the host where the flotilla is physically connected.

I’d wholly agree with using the websocket interface, since it will be the defacto way of interacting with Flotilla across what will become, no doubt, a range of services.

I’m in the midst of a complete overhaul of the websocket Daemon, Dock firmware and Rockpool at the moment which will make Flotilla generally much faster and more responsive- it should rock for node-red as much as it does for Rockpool users. As part of the update it will also include a Discovery service ( @Jon’s excellent idea )- flotillad will publish its LAN IP Address to our remote discovery server, saved against your WAN Address, you can then use node-red to query the server and get a list of local Flotilla daemon’s back in json.

Most of the software is kept in a rolling usable state with intermittent build breaks- so if you want to have a play with pre-release dock firmware and daemon to see how it feels in node-red, hit me up with a PM and I’ll get some instructions organised.