Real-time Mobile Data Analysis Using Watson Analytics | Digital Conqueror

Real-time Mobile Data Analysis Using Watson Analytics

Based on machine learning this tool takes advantage of its deep learning to understand user behaviour, decision making and do the unbiased analysis exactly similar to human brain, and create a prediction that can give businesses deeper understanding of their customers for taking better decisions when it comes to making products, offers, and where to focus when it comes to marketing.

Watson Analytics also unifies technologies like the natural language processing, semantic recognition, and dynamic starting points and discoveries in a single tool. The incredibly simplified dashboard needs no special skill sets like the programming, statistics or data science.  The heavy lifting is completely done in the background, and you can take full advantage of the analytics platform no matter if you are a sales person exploring the performance of the last quarter or a well-trained data scientist training some preliminary models.

This cloud-based analytics tool enables you to naturally dialogue with your data and reports and primarily designed to work with structured data. I am going to explore more about how we can interact with data sets in upcoming articles in this series, but today we are taking a look at interesting real-time data analytics capabilities using IBM Bluemix and my smartphone.

We’re using my smartphone as an IoT device, that we will connect with the Watson IoT to get real-time smartphone movement data to Watson IoT application. With this data, we can create automated rules and actions like sending email alerts or creating applets with IFTTT.

This is a beginner level tutorial, and even I am trying this for the first time, so you can guess it’s quite easy to follow.

Step by step Mobile Data Analysis in Real-Time with Watson Analytics

1. Signup & Login to IBM Bluemix Platform.

2. We basically want a setup where a smartphone is connected to Watson IoT service in the Bluemix platform, and we are using a pre-built solution available here-

3. Click on the Create Toolchain button to start the deployment.

4. Create your organization and select country. Name your space dev or test.

5. You can keep the default App Name for simplicity, but you can customize it. Make sure you note the name and hit the DEPLOY

6. Now click on the Delivery Pipeline to initiate and verify your app passes the cloning the build and deployment stages, which it generally does. This can take 3 to 5 minutes, and you will be notified once the deployment completes.

7. Now to see the URL of your application, click on Menu button on the top left > Apps > Dashboard to visit the dashboard. Here you will see the Route/ URL of your application.

8. Use this URL on your smartphone browser. And you will see this prompt. I have added a unique device ID oneplusx-01.

The device is now communicating with our app deployed on the Bluemix and delivering the acceleration and movement data in real-time.

NOTE: Make sure device display is not locked, and the location is turned on to keep sending the data.

9. Now head over to iot-phone-iotf-service, then hit the Launch button.


10. On the left menu, go to Devices and you will see your device ID with signal bars symbol that ensures your device is connected.

11. By clicking the device id > recent events you can see the event data is displayed in real-time.

Creating Schema for our IoT Device (Smartphone)

To take advantage of Watson IoT platform and apply rules or actions, we must implement the Schema that will help us name the device properties to readable/user-friendly property names.

1. Now click on Manage Schema > Add Schema and select your smartphone from the dropdown.

2. Click on Add Property > select From Connected > select All and click OK & Click Finish to complete this process.

Creating Rules & Actions

Now we want to move to Create Action phase of the Watson IoT Cloud Analytics, where you can trigger particular action based on the device data. You can also setup rules like ‘Displaying alerts on Watson Dashboard, Sending an email or even trigger a particular webhook’ if certain device data matches the number of events. Let’s see how that works in real-time.

1. On the left side menu go to RULES >Click on Actions > Create an Action

2. Fill up the action details and finish creating the action. You will see the action created in the table.

3. Now click on browse > Create Cloud Rule

4. We are triggering the rule if Beta value (ob) touches 100 or more , and selecting our previously created action for the same. Make sure you write the proper description for further analysis and alterations. You can also use conditions like AND or OR to create the deep filtering of data with the available properties.

5. Now click on blue link ‘Trigger every time condition is met’ and select the intervals you want to get alerted based on your task priorities.

And once you set all these parameters you can click Activate to launch the analysis and monitoring sequence.



Analysis & Results

1. Head over to the boards on the left menu to see the results.

2. Device Centric Analysis card will show you alerts for the particular devices as follows. I have gone ahead and added few more rules, which you can see as the ‘Alpha Threshold’ data.

So this is how simple it is to see real-time data from your IoT device or Smartphone using Watson IoT Cloud Analytics. Possibilities are limitless with rules you can connect with IFTTT applets or web hooks to develop data-rich mobile or web applications.

In next the article, we are going in depth with Watson Data Analytics platform, and how you can leverage it to build and analyze predictive models, visual data representation, and asking questions to Watson platform with Natural Language Programming.

Amol Wagh

Author of Game Marketing Book: Market Your Indie Game Like A Pro! 7 years of Tech Blogging & Digital Marketing. Co-founder at Dotline Inc & a Gamer. Follow me on Twitter @amolwagh, @amol.wagh


No comments!

There are no comments yet, but you can be first to comment this article.

Leave reply