WIA IoT Platform

In this tutorial you will use Wia.io to collect and display data from the RPi. You will also use Wia to send commands to a device.

Set up

Create an account on Wia.io
On your Raspberry Pi, install Wia by opening a terminal window and running the following command:

pip install wia

Make a directory called iot-week9 for your python programs

mkdir iot-week9

Create a Device in Wia

Go to the Wia Dashboard and select Create a New Space then select Devices. Add a device and give it the name SensePi. Now, in the Configuration tab for your device, you will find device_secret_key which should begin with d_sk. This will be important later on.

Python Code

In the iot-week9 directory, create a file called sensehat_wia.py containing following code:

from wia import Wia

wia = Wia()
wia.access_token = "Your access token"

wia.Event.publish(name="temperature", data=21.5)

Run the program. In the weather space on the Wia Dashboard, select Devices and check the temperature event has appeared in the Events tab for your device.

Temperature Event

Select the Overview tab and click the Add a Widget button. Add a widget called Temperature. For the event field, make sure you type the event name exactly as it appears in the code (mind your case!). Your overview tab should be similar to the following:

All going well, you now have code that interacts and creates events in Wia

Integrate SenseHat

Now lets update the code to use the SenseHat sensor values to create events:

Replace the code in senshat_wia.py with the following code:

from wia import Wia
from sense_hat import SenseHat
sense = SenseHat()

wia = Wia()
wia.access_token = "Your access token"

temp=round(sense.get_temperature(),2)
wia.Event.publish(name="temperature", data=temp)

You are now taking the temperature sensor reading from the SenseHat and publishing it to Wia

Now, as before, run the script again and check your Wia space responds to the event.

Exercise

update sensehat_wia.py to do the following:
- to create pressure and humidity events in your Wia space every 15 seconds.
- Add text widgets for pressure and humidity.
- Change the Temperature widget type to a graph(leave default values for Time period and Aggregate function) All going well, your overview tab should now look like this and update every 15 seconds.

Wia Overview Tab

Web Page

In the same directory as your python script, create a new directory called html.
Create a file called index.html and add the following content:

  <!DOCTYPE html>
<html>
  <head>
    <meta charset="UTF-8">
  </head>
  <body>

    <h1>SenseHat Data</h1>

  </body>
</html>

Navigate back to your Wia Dashboard. In the overview for your device, you can see your widgets. In the upper right hand corner of the widget, there should be a box with an arrow. Click the box. A screen like this should pop up.

Select Anyone can view this widget and embed it in any website. You should also see Embed code, which will start with <iframe> and end with </iframe>. Copy the entire code and paste it below the <h1>SenseHat Weather Station</h1> line and above the </body> line.
View your index.html page in a browser. It should look similar to the following:

HTML Page

Put it on the web

You can use GitHub to host your webpage so that anybody on the web can view it.

If you don't have a github account already, you can make one here.

Once you are set up with github, create a new repository and name it your-github-username.github.io. Check the box to initialize with a README.

Now, navigate to your new repository and create a new file. It must be named index.html. Copy and paste the code from index.html. Click commit changes. Now, visit your site at https://username.github.io. You're on the Web!

Smile, you're on camera.

You will now use Wia events, commands and flows to control the SenseHat using facial expressions. Wia uses MQTT and the publish-subscribe pattern we talked about in class in their commands functionality.

Create Commands

Inside your Device Dashboard click on the Commands and then click Add Command
Add a happy-face and sad-face commands. You will use these commands to control the Raspberry Pi with a smile!

Add Photo Command

When finished, the commands tab should look like this:

Commands

Create Photo Events

You will now write a small program that will take a photo and create photo events in Wia. PLEASE CHOOSE ONE OF THE FOLLOWING OPTIONS TO ACCOMPISH THIS

OPTION 1: Using a Raspberry Pi Camera

Connect a Pi camera module to the Pi. Follow the instuctions here
Open a terminal window on the RPi and enter sudo raspi-config
Select option 5 Interfacing Options

raspi-conf

Select P1 Camera and enable the camera interface:

Enable camera

Your camera is ready to go! Exit raspi-conf by selecting back/exit.

Create Python Program

In the directory you created earlier called iot-week9 create a file called snap.py.
Enter the following code in the file and save.

from wia import Wia
import time
from picamera import PiCamera

wia = Wia()
## INSERT YOUR SECRET KET
wia.access_token = 'YOUR_SECRET_KEY'
camera = PiCamera()

## Halt execution until
input('Look at the camera and hit "Enter" to take a pic...')

## Start up PiCam
camera.start_preview()
## sleep for a few seconds to let camera focus/adjust to light
time.sleep(5)
## Capture photo
camera.capture('/home/pi/image.jpg')
## Stop the PiCam
camera.stop_preview()

## Publish "photo" event to Wia. Include the photo file. 
result = wia.Event.publish(name='photo', file=open('/home/pi/image.jpg', 'rb'))

Now, open a terminal window in the iot-week9 directory run the script by entering the following command:

python3 snap.py

You should see an image file appear in /home/pi/image.jpg.
Go to Wia and check that a new photo event has appeared.

Photo Event

OPTION 2: Using a Laptop Webcam

Post photo from webcam
Install OpenCV

sudo pip install opencv-python

If not already installed, install Wia

sudo pip install wia

On your laptop workstation, create a directory called python-photo.
Create a file called snap.py and enter the following code:

import cv2
from wia import Wia
import os
import time

input('Hit any key to take a pic...')
vc = cv2.VideoCapture(0)
wia = Wia()
wia.access_token = 'YOUR_DEVICE_SECRET_KEY'
file_name='wia-pic.jpg'

if vc.isOpened(): # try to get the first frame
        rval, frame = vc.read()
        cv2.imwrite(file_name,frame) # writes image test.bmp to disk
        dir_path = os.path.dirname(os.path.realpath(__file__))
        result = wia.Event.publish(name='photo', file=open(dir_path + '/' + file_name, 'rb'))
else:
        rval = False

Change the wia.access_token to your key and run the program.
Look into the camera and press andy key. Your photo will be published as an event on Wia. Log into Wia and you should see the event recorded, similar to the following.

The photo event

Wia Flows

You will now create a Flow that is triggered by photo events:

In the Wia dashboard, click on the flow option on the left side. Create a new Flow as follows:

Photo processing flow

The trigger is when a photo event is created by the sensepi device.
This then goes through a Detect Faces service node, the output of which branches off into two Run Function logic nodes; one to output a string "Yes" if the subject is smiling, and one to output a string "No" if the subject isn't smiling. Here's the code for the 'smiling' logic node:

if (input.body.faceDetails && input.body.faceDetails.length > 0) {
   output.body.isSmiling = input.body.faceDetails[0].smile.value;
  if (output.body.isSmiling){
    output.process = true;
    output.body.data = "Yes";
    }else{
    output.process = false;
    }
} else {
    output.process = false;
  output.body.data = false;
}

The Javascript code for the not smiling node is as follows:

if (input.body.faceDetails && input.body.faceDetails.length > 0) {
   output.body.isSmiling = input.body.faceDetails[0].smile.value;
  if (!output.body.isSmiling){
    output.process = true;
    output.body.data = "No";
    }else{
    output.process = false;
    }
} else {
    output.process = false;
  output.body.data = false;
}

If the subject is smiling, the 'happy-face' Command is run, triggering the RPi to display a happy emoticon on the SenseHat. If the subject isn't smiling, the 'sad-face' Command is run, displaying a sad emoticon on the SenseHat.

Now add a photo widget on the SensePi devices overview page and link it to the photo event as follows:

As an exercise, add a Text widget and link it to the happy event.

Update Raspberry Pi

Open a terminal window and run the following command in the iot_week9 directory containing sensehat-wia.py

wget http://rpf.io/shfaces -O faces.py

Update the code in sensehat_wia.py to subscribe to the commands and show the corresponding emoticon by updating the code to the following:

from wia import Wia
from sense_hat import SenseHat
import time
from faces import normal, happy, sad

# happy face callback
def on_happy_face(event):
    print(":)")
    sense.set_pixels(happy)

# sad face callback
def on_sad_face(event):
    print(":(")
    sense.set_pixels(sad)

sense = SenseHat()

wia = Wia()
wia.access_token = 'd_sk_JFODUgcmYToVZSd7JP8xrt54'
deviceId = 'dev_NEJjk3oa'

wia.Stream.connect()

# Subscribe to happy and sad face commands
wia.Command.subscribe(**{"device": deviceId, "slug": 'happy-face', "func": on_happy_face})
wia.Command.subscribe(**{"device": deviceId, "slug": 'sad-face', "func": on_sad_face})


while True:
        temp=round(sense.get_temperature(),2)
        press=round(sense.get_pressure(),2)
        hum=round(sense.get_humidity(),2)
        #publish temp/pressure/hum
        wia.Event.publish(name="temperature", data=temp)
        wia.Event.publish(name="pressure", data=press)
        wia.Event.publish(name="humidity", data=hum)    
        time.sleep(60)

Run the program on the RPi and take another photo. It should now indicate if you're smiling or not. Smile! You shoud see the corresponding result on the RPi Sensehat.

Happy Pi