Skip to content

Latest commit

 

History

History
151 lines (108 loc) · 8.96 KB

Lab2_SmartLampVisualize.md

File metadata and controls

151 lines (108 loc) · 8.96 KB
typora-copy-images-to
images_Lab2

Lab 2: Smart Lamp data visualization with Watson IoT Platform on IBM Cloud

Introduction

In this lab, we will use two different techniques to visualize the IoT sensor data collected on the Watson IoT platform.

We will first use the real-time insights dashboard provided by the platform, and then use an external application, in this case Node-RED running as a Cloud Foundry application on the IBM Cloud.

Prerequisite

This lab assumes that the RaspiLamp types, schemas and devices of Lab 1 have been setup.

1. Creating a dashboard on Watson IoT Platform

We will now create a graphical interface to plot the data of the various sensors of the lamp.

  • Select Dashboard menu

  • Click on the big empty rectangle to add a table

  • remplir les données
    then Next and Submit

  • Open the newly created table:

  • Click Add New Card to add items to the table

Adding a gauge

  • For example, add a brightness gauge
  • Check that the type is Number
  • Choose size and design of your gauge:

Adding a multi-value graph

  • Choose a card Line chart:
  • Add two datasets, one for ldr:
  • Another on for solar:
  • Choose a size chart L:
  • The values of the two sensors will be displayed on the same diagram

You can create maps for other sensor values.

2. Creating an interactive dashboard with Node-RED

We will now create a dashboard to act on the lamp. For this we will use Node-RED, but directly out of the Cloud.

Starting Node-RED

  • Return to the IBM Cloud Dashboard (https://console.bluemix.net/dashboard/apps)
  • You must have an application running;Select it
  • In the Routes menu select the first one :
  • The node-RED setup screen is displayed, click Next then forcez la configuration non-sécurisée
  • Press Next and then Finish
  • Finally, click , which opens the Node-RED flow editor.

Creating a "send command" flow

  • Flow 1 predefined does not interest us necessarily, we will create another thanks to the + top right;

  • From the palette on the left side of the screen, add a ibmiot send node . Warning, there are two similar nodes, this one has an entry on the left, and it comes from the drawer output

  • Add an Inject node and wire them:

  • modify the properties of the injected node to send a string of characters: like RGB for instance

  • Changeibmiot node parameters to match your lamp:

    • Authentication should be Bluemix Service
    • Output Type must be **Device Command**
    • Format must be text,

  • Activation of the entry (left pellet of the node) of the inject node should turn on your lamp.

  • Note that this interface is accessible from the internet now, which was not the case in the first part.

Adding a dashboard

Setup

  • From the menu at the top right, select Manage Palette:

  • Toggle on the tab Install

  • Type node-red-dash in the search field

  • Click on the Install button This added a series of nodes in the left palette, under the category Dashboard:

Creating an interactive command dashboard

  • From the palette, add a Text input node and wire it to IBM iot:

  • Change node settings: Create a group:

  • Then a tab

  • Accept defaults or change names to your liking

  • Validate and deploy, then click on the table launch from the new tab dashboard on the top right:

  • The dashboard is displayed:

  • You can now interactively change the color of the lamp by typing for example W or a sequence

Adding gauges

We can decorate this dashboard of various displays:

  • Add 3 nodes from the palette:

    • ibmiot with do on right
    • change
    • gauge

  • Configure the nodes as follows:

    • ibm IoT, modifyAuthentication:
    • The change node is used to extract the incoming message data to the gauge node format, we extract here the value of ldr:
    • For the gauge, we will change the display limit:

  • Wire the nodes together: then deplo

  • When we go back to the dashboard, we now see a gauge that changes when the sensor is more or less lit:

This interface is now exposed directly on the internet, you can connect to its public URL from a smartphone for example.

Be creative!

You can change the dashboard at your convenience!

3. Control and automation of the SmartLamp

Last step, we will program a simple automation mechanism that changes the color of the lamp according to the ambient light level.

Color change based on ambient light

  • Add a switch node and set it with the button Add as follow:
  • The node now has 3 outputs. Add 3 change nodes , and configure each one for a given R G B color:
  • Wire as follow and deploy

The lamp will change color depending on the illumination of the LDR sensor.

4. Lab Stretch Goal

For this section of the Lab, you will be tasked with a creating a more complex Dashboard, without precise instructions but just a functional and visual description, and a few hints.

The goal now will be to create a more elaborate dashboard, which will look like the following:

  • The line chart graph shows the value of both the solar and ldr sensors
  • The set of buttons on the left will send a single W,R,G,B,0 command to the RaspiLamp
  • The text entry below will send a command string when the [Set] button is clicked.
  • The Clear button will reset the line graph
  • The 4 dial gauges below will reflect the current value of the LEDs.

Implementation Hints:

  • To set multiple values on the line graph, each sensor value must be passed in msg.payload with a msg.topic set to the sensor ID. You can use a change node for that, e.g.:
  • The LEDs colors are in the d.leds, as a WRGB field, with an octet for each element of the LED. You will need a function node with multiple outputs to split the LED colors into 4 outputs:
  • To clear the line graph, you must send it a msg.payload with an empty array:

Be creative!

On this model, you can design different dashboards or automations loops variations to your liking!

END of Lab 2