Do you want to be able to bring one of your toy lightsabers to life using nothing but your voice? Then this is the project for you! Using an Echo dot, any computer that can run Python and a Walabot, we're going to add special FX and voice control.
DemoFirst StepsFirstly you will need a Walabot, and Amazon Echo / Dot, a computer capable of running Python and a toy lightsabre. Got that lot? Good!
Initial setupTo run this program properly you will need to make sure you install some frameworks to help us communicate with Alexa.
Firstly, from the command line type:
pip install flask
This will install the flask framework modules for Python to use. Next:
pip intall flask_ask
Flask ask is very nifty and runs a server locally that will interface with the Amazon Cloud. The only issue is: how do we get the Amazon Cloud to see it if it's only available locally? Ngrok to the rescue!
Go to ngrok sign up (for free) and download it for your OS. Follow the excellent instructions on the website to get it configured and then from the location of the downloaded ngrok executable, type:
ngrok http 5000
This will start a service that will open a tunnel from port 5000 (this is what flask_ask will be using) to the cloud via a randomly assigned URL. You will need the https URL it gives you, take a note of it. We will need this when setting up the Alexa skill.
WalabotWhat is a Walabot?
Want to see through walls? Sense objects in 3D space? Sense if you are breathing from across the room?
Well, you're in luck:
The Walabot is whole new way for sensing the space around you using low power radar.
Head on over to api.walabot.com and download the latest walabot drivers for your OS. make sure you follow the python installation instructions. You'll need to python walabot wrapper for this to work.
For this project I used the Beta version of the SDK, I was eager to try out the new GetTrackerTargets functionality. This makes target acquisition and tracking way more accurate. You can download the installer for Windows and RPi here:
https://walabot.com/getting-started-beta
This installs everything you need to interface with the Walabot.
The CodeThe best thing to do is to either download or fork my github repo for this project This can be found here: https://github.com/daveyclk/Alexa-Lightsabre-Trainer
In here you will see a few separate sheets of code. This keeps things nice and neat. It is very easily stitched together. The sheets are:
- main.py - This is the program you will run. This sets up threads for the various servers and programs running.
- config.py - this contains the variables that are shared between the sheets
- lightsabre.py - This controls the UI, walabot tracking ans special FX sounds
- lightsabreAlexa.py - This is the flask ask app that will communitcate with the alexa service
- templates.yaml - This holds all the phrases Alexa will say. These are rendered within lightsabreAlexa.py
Looks scary, but it isn't! Provided you have followed the initial setup steps, all you need to do is type this into your command line (this is a windows machine):
python main.py
You will see the startup screen pop up:
You will also see the command prompt showing the flask_ask app server running on 127.0.0.1:5000 this is the reason we needed to point ngrok to port 5000.
Thats it for the app... but until we add Alexa support, it will do nothing.
Alexa SkillNow wait a minute. If you think for a second this is going to be hard, guess again. Amazon have made skill creation very VERY easy. This is the first skill I have created, and this method made it completely pain free. Firstly you'll need to sign up for an amazon developer account here.
Once you've been fully authorised (this can take a day or so as I found out) you're ready to go. Next you'll want to begin to create your new skill go here to begin: https://developer.amazon.com/edw/home.html#/skill/create/
For this skill we only need to work about the first 5 tabs. This is the Skill information. This is where you fill out the basic details. You need to give it a name and also an invocation name (this is what you shout out to start it up)
The Alexa skill ID for my skill is - amzn1.ask.skill.3c851c50-cab7-452e-a148-10fb3c822b56
In this case the name and invocation name are the same. Light Saber Trainer:
The next page is the interaction model. This is the part where you tell the Alexa service what is going to be said and what sort of responses she can expect:
For the purpose of this tutorial, just copy and paste the code from my github:
- /alexa/lightsabre_intent.json shoud be copied into the intent schema
- /alexa/sample_utterances.txt should be copied into the sample utterances
The next page is the configuration. This is the part that points the Alexa service to our project. Remember the ngrok url you created? You need that now. Select https and in the default field, paste the ngrok URL. Copy all the other radio buttons:
Second to last is the SSL certification page. as we're using ngrok, you need to select the second one in the list:
And lastly we're going to look at the test page. This is where you enable you skill for testing on your Dot/echo. But you can also test it using their beta test service. It's all very handy!
Once you have completed all those things, you should be in a position to have a play! Set up the Walabot as shown in the video, you'll want to stand about 1.5 meters away to the side.
To start the app, simply say "Alexa, start Light Sabre Trainer". She will ask you if you'd like to open your light saber, you can say yes at this point.
If you're out of the set arenas range, the light saber will look Half Bright. Like this:
When you move closer it will look like this:
Start working on your best light saber moves! You will see various lightsabers hit images:
At anypoint you can say "Alexa, start Light Sabre Trainer".
You have a few options, you can ask Alexa what your score is, or you could say finish or you're tired and Alexa will respond with either a current hit score or she will close the light saber and the FX will stop.
CaveatsI am still working on refining the Alexa utterances and general feel of her personality. The arena area of the walabot may need to be tweaked depending on your environment. You can do this my modifying the this code in lightsabre.py:
R_MIN, R_MAX, R_RES = 10, 120, 10 # SetArenaR values
THETA_MIN, THETA_MAX, THETA_RES = -20,20, 10 # SetArenaTheta values
PHI_MIN, PHI_MAX, PHI_RES = -45, 45, 10 # SetArenaPhi values
TSHLD = 1 # SetThreshold value
Important!Please enjoy responsibly and let your children have a go.
Comments