Process and Visualize Historical IoT Data Using Azure Time Series Insights

45 minutes
  • 4 Learning Objectives

About this Hands-on Lab

In this lab, you will explore a Time Series Insights Gen2 (TSI) demo site that Microsoft has set up and loaded with a large dataset, which has enough variety to allow you to more meaningfully explore this data analytics tool. While using the provided lab diagram as map of the user interface (UI), you will get practice using the TSI explorer, while collecting intelligence about the specific dataset loaded into the demo.

Learning Objectives

Successfully complete this lab by achieving the following learning objectives:

Explore TSI Features
  1. Log in to the Azure portal in an InPrivate or incognito window, using the credentials displayed on the lab page. Note that you do not need to use the portal for this lab. We’ve spun up the subscription, just so you have a valid login for the Time Series Insights demo site.
  2. Navigate to the Time Series Insights Gen2 demo, using the link supplied in the Additional Resources section of this lab. You should be logged in automatically, but if prompted, use the same credentials you used to log in to the Azure portal.
  3. Retrieve the lab diagram from the lab page, and use it as a guide for the following steps.
  4. Change from the current Analysis mode to the Model mode. Change back to the Analysis mode.
  5. Change the timeframe you want to display by selecting a region on the chart, using the slider, or manually selecting specific dates from the date range displayed. Any timeframe is fine.
  6. Determine how to smooth the current lines in the chart by modifying the chart interval. (This isn’t amplified in the lab diagram; it’s your job to find it.)
  7. Navigate the Contoso Windfarm Hierarchy to find the WindDirection sensor type for the W6 windmill at Plant 1. Select the instance (time series), under that sensor, and add the Reading to the chart and the time series well.
  8. On the chart, right-click on a line to toggle between showing or hiding min/max shadows.
  9. Drag over a section of the chart, and right-click to Explore Events and then View Statistics.
  10. Explore the information available in the time series well.
  11. Clear all instances from the time series well (and, therefore, the chart).
Determine Consistency of Active Power Across Windmills
  1. Find the ActivePower time series instance for all 4 of the windmills, and add the value for each to the chart.
  2. Shorten and lengthen the time span to answer this question: Is the ActivePower measure consistent across all 4 windmills in the majority of the observable time spans?
  3. After answering the question of consistency, remove all instances from the chart except for ContosoFarm1W6_GenPower1. Leave that one on the chart.
Confirm that Wind Speed and Active Power Are Correlated

Common sense tells us that, generally speaking, wind speed and windmill performance are directly correlated. However, we should confirm this understanding to ensure the dataset appears to be delivering valid values.

  1. On the W6 windmill of Contoso Plant 1, find the WindSpeed instance, and add the Reading to the chart.
  2. Zoom and pan and adjust the interval to try to answer this question: Is the Active Power measure correlated with wind speed?
  3. Once you have answered this question, remove the WindSpeed instance from the chart.
Determine If Temperature and Active Power Are Correlated

Not being wind power experts, we probably don’t know if outside temperature is correlated with active power. Let’s ask the data.

  1. On the W6 windmill of Contoso Plant 1, find the OutdoorTemperature instance, and add the Time Weighted Reading to the chart.
  2. Zoom and pan and adjust the interval to answer this question: Is the ActivePower measure correlated with outside temperature? Does switching from the Time Weighted Reading to the Reading change your answer?

Additional Resources

Imagine you are a new data engineer at Contoso, Microsoft's favorite all-in-Microsoft customer. You are part of a team building some machine learning solutions over IoT data sourced from Contoso's wind farm. To assist the analytics team with the initial build of machine learning models, you have been tasked with examining and characterizing the shape of the existing data in order to answer fundamental questions around the relative performance of individual windmills and any correlations between performance and the readings coming from other sensors. The required dataset for this analysis is loaded into Time Series Insights, or TSI. You have never used TSI before, but you are up for the challenge. And because you're new at Contoso, you have no meetings on your calendar. You have plenty of time to explore and ponder.

Accessing the TSI Demo Site

Use the following link to access Microsoft's Time Series Insights Gen2 demo site.

If you open the demo site in the same browser window you used to log in to the lab, you should be logged in to the demo automatically. If not, and you are prompted to provide a username and password, use the login credentials from the lab.

WARNING: Be Prepared for UI Changes

Given the fluid nature of Microsoft cloud tools and open-source projects, you may experience user interface (UI) changes that were made following the development of this hands-on lab that do not match up with lab instructions. When any such changes are brought to our attention, we will attempt to update the content accordingly. However, if changes occur, students will have to adapt to the changes and work through them in the hands-on labs as needed.

What are Hands-on Labs

Hands-on Labs are real environments created by industry experts to help you learn. These environments help you gain knowledge and experience, practice without compromising your system, test without risk, destroy without fear, and let you learn from your mistakes. Hands-on Labs: practice your skills before delivering in the real world.

Sign In
Welcome Back!

Psst…this one if you’ve been moved to ACG!

Get Started
Who’s going to be learning?