Vivid 2018 LED IoT Apparel !

Hello! 

Continuing with the tradition of creating something beautiful and covered in LED lights, this year we have something special!

Previously, we had the LED Jacket with Tearschu, and the LED Dress with Naifel. Taking inspiration from these, and solving a lot of the problems I faced with them, I bring the latest iteration of light up fun.

This year, I have taken a pair of high heel boots, and an umbrella from Daiso, added plenty of pretty lights, and of course, this year marks the first year the entire project is connected to the internet.

 

 

The project was built using mostly the same core components for each item. 

The shoes each have:

  • Lithium Ion Battery (1000mAh)
  • LiIon Charge / Boost circuit MP2636 
  • WeMos D1 Mini
  • A random switch for power
  • A strip of WS2812 LEDs

The umbrella is similar, except instead of the MP2636 boost circuit and 3.7v Lithium battery, I used a 3s LiPo battery, and a 5v step down regulator capable of high current.

 

The physical build was pretty straight forward, hook up everything how you please, battery to boost/charge, from there to the WeMos / LEDs, and then route the wires how you please. For LED placement on the shoes, I went with up along the front as I feel this will look the best having the light cover the most area, and for the umbrella I ran the lights down the spokes of the umbrella.

Unfortunately with my design you cant really CLOSE the umbrella anymore but as this is just for Vivid I am not too fussed 🙂 

To attach the LEDs to the umbrella I initially tried to use hot glue, but it was actually melting through the umbrella, and the parts that didn’t, did not hold very well, so I ended up using clear packing tape, as it does not seem to get in the way of anything and is barely noticeable! 

The LEDs here are hooked up in parallel with each other, so each spoke on the umbrella will be the same.

This slideshow requires JavaScript.

Once it is all made up physically, we can move on to the code.

I was looking into using the McLighting project for control of these, as it has both an internal web interface as well as support for things like MQTT, but I could not get it to work reliably, and it didn’t support running in AP mode, only client mode, which was a big turn off for me.

So what I ended up using was the JSON LED code from BRUH Automation, because I use this for other things at home and it works pretty reliably.

One thing to note here, for my LEDs I had to add the following two lines of code, BEFORE including the libraries, to prevent flickering of the strip. (not sure why this works?)

#define FASTLED_INTERRUPT_RETRY_COUNT 0
#define FASTLED_ALLOW_INTERRUPTS 0

 

(at the verryy top of the sketch)

 

Now my initial plan included taking a small portable router, and a Raspberry Pi 3 out with me to vivid, running a local MQTT server on a local network, with the Pi running Home Assistant (Hass.IO) all locally so I could connect to it to control things. However I ran into many problems attempting to do this, I am not sure if its because I don’t know how to properly setup static IP’s in resin, or just because it hates me, but I kept not being able to connect or it wouldn’t respond to my commands, it just wasn’t working great.

 

One day though, my good friend Mark came over and we needed a project to work on, so what we set up was a private mosquitto MQTT broker, that requires authentication, running in Docker on a Ubuntu Server 18.04 LTS install!

What this meant,  was I now had a secure way of connecting a remote node to my Home Assistant running back home.

 

I went right ahead and adjusted the code on the three items for the new server, forwarded the ports in my router, added the config to my production Home Assistant server, and hey presto, was I glad to see, everything JUST WORKED.

 

I made a view in Home Assistant and threw all the entities into it, and here’s how that looks:

So as you can see, we can control both shoes together, each individually, the umbrella on its own, or everything as a group!

We also can change the animation speed of the various effects.

I will be heading out to vivid to shoot a small video and some photos with this, with my good friend Tsugumi modelling it for me, on the 9th of June 2018 from about 6PM onward, Not sure if I will be at Circular Quay or Darling Harbor yet, keep your eye on my Instagram to find out! 😉 

 

Thanks for reading!!!

Image Recognition for Home Assistant

I was browsing the Home Assistant Community Forums earlier today, when I noticed a post by Robin Cole!

Rob has created a custom component for Home Assistant that allows us to use image classification via Tagbox, on a camera feed that Home Assistant has access to.

Following the instructions on github was pretty straight forward, the only thing I changed was to reference an external image_procesing.yaml rather than having it all in my configuration.yaml as I’m trying to be a bit cleaner.

 

In this case, my docker is running on Windows, the Tagbox instructions were fine for this, I didn’t realise you can just run the image and if its not installed, it will download it 🙂 

 

So now that it’s all setup it appears to be working great! 

As you can see, the component creates a new entity called image_processing.tagbox_name, with a state of the most likely item in the image, and then some attributes of the next most likely, and a count of any specific tags from the config.

I’ll play around with it to see what I can use it for more practically,  I can think of all sorts of things this component will be useful for! From identifying if a car spot is free or taken, to seeing if the bins were already taken out! 

 

I just wish I could specify a region within a camera feed for it to analyse.