Enabling Enterprise-wide Data Science with OCS Dataviews

About this course

In this course, you will gain hands-on experience with a new OSIsoft software product, OSIsoft Cloud Services (OCS). OCS is a cloud-native platform built for historical, real-time and future/forecasted operational data. OCS complements existing on-premises PI systems and enables users to easily define, visualize, query and shape data sets required for data science.

By the end of the course, you will be able to...

Use the features and functionalities in OCS
Analyze & visualize data in OCS
Add metadata & create Data Views in OCS
Access OCS data in a Jupyter Notebook via Python
Filter & prepare data from OCS for machine learning
Use a machine learning model on a filtered OCS dataset
Retrieve weather data from a Weather API into the trained machine learning model
Send forecasted data back to OCS

Audience

This course is primarily suited for data scientists who are looking to extend operational data in their PI Systems for enterprise-wide machine learning applications.

Level: Intermediate

Study time: 16 hours

Course Access: Unlimited access to all content except the Training Cloud Environment (TCE). You have 30 days access to the TCE starting on the day you access module section "Launch Cloud Environment".

After those 30 days you can purchase additional access with one of the two options below:

Prerequisites

Basic understanding of PI Server components, including PI Data Archive and PI Asset Framework
Basic understanding of Python, Jupyter Notebooks and machine learning
A computer that can access our YouTube content, and pass our connection test

Software Requirements

In this course, we will work on a virtual machine hosted in the cloud, which contains the following components:
- Access to Active Directory
- PI Data Archive 2017 (3.4.415 or later) with PI System Management Tool (PI SMT)
- PI AF Server 2017 or later with PI System Explorer
- PI Vision 2019
- Access to the OSIsoft Cloud Services Namespace "Enabling Enterprise-Wide Data Science" using a Microsoft or Google email account

Microsoft Office or equivalent is required to access the workbook and presentation provided.

This Course Includes...

Videos, discussion opportunities and quizzes to help you learn the material
This course is a hybrid self-paced/facilitated course for your convenience. The course is designed to be done without an instructor; however, there will be a facilitator available for the first week of the class to make sure that you get the necessary access to your course specific virtual machines and the permissions to interact with OSIsoft Cloud Services technologies.
Please use the video lectures for instruction along with the course exercises to gain experience working with the key concepts presented.
There is a final quiz which you must pass to obtain a certificate of completion.
Once you register for a course, you will have access to the course materials 24/7 on this website.
Please note: access to your course virtual machine will only last for the first 30 days of the course (Day 1 is the first day of your facilitated week).

Further Information

This course assumes basic knowledge and understanding of wind turbines and their corresponding power curves. A brief introduction of wind turbine operation and relevant aspects of their power curves will be presented, but for additional information, please refer to the References section at the end.
This is a self-paced course. Any questions or assistance needed about the material can be asked in this course's space in the OSIsoft PI Square community
When you complete the examination at the end of the course, you will receive a certificate of completion which can be shared and directly posted on LinkedIn.
For more information about our Online Courses please visit our FAQ page

Course Material

You can audit the full video lecture content right now on the OSIsoft Learning YouTube Channel
Download Course Workbook.

Curriculum

Getting Started
PLEASE READ - Key Course Information
Cloud Services Invitation Information
How to Navigate This Course
Discussion Forum
Offline Course Videos for Blocked YouTube Users
Course Workbook
Training Cloud Environment
Launch Cloud Environment
Lesson 1 - Introduction
Introduction [3:31]
Lesson 2 - Background Knowledge
Domain Knowledge – Machine Learning [5:28]
Review the Wind Farm AF Model in PI System Explorer [1:58]
Lesson 3 - Introducing and Exploring OCS
Introducing OCS [2:17]
Logging into & Navigating OCS [12:44]
Lesson 4 - Managing metadata in OCS
Metadata in OCS [8:15]
Lesson 5 - Creating Data Views in OCS
OCS Data Views [5:25]
Lesson 6 - OCS clients
OCS Clients [3:39]
Lesson 7 - Accessing Jupyter Notebook & Using API Console in OCS
How to access Jupyter Notebook [5:50]
Introduction to OCS API Console and documentation [1:45]
Prepare and check URL calls to access data in OCS [4:07]
Getting data from data views [5:55]
Lesson 8 - Accessing OCS Data View and Preparing Dataset for Machine Learning
Introduction [1:35]
Modify config.ini file and create new secret [7:34]
Wind Turbine script walkthrough – part 1 [10:57]
Wind Turbine script walkthrough – part 2 [9:13]
Create new data view for turbine states [5:32]
Wind Turbine script walkthrough – part 3 (Filters) [3:09]
Lesson 9 - Use Filtered Dataset to Train & Evaluate Machine Learning Model
Introduction [2:22]
Prepare the training and testing data sets [4:21]
Use the Decision Tree Regression ML model [2:12]
Plot the results and save the ML model [3:09]
Test the model and run sample prediction [5:18]
Lesson 10 - Getting Forecast Weather Data into ML Model
Introduction [2:02]
Retrieve weather forecast [2:52]
Store JSON response in data frame [2:10]
Use the ML model to predict Active Power [2:39]
Lesson 11 - Sending Data Back to OCS
Introduction [1:48]
OMF overview [6:40]
OMF and connections [1:48]
Send data to OCS (create SDS Type) [7:23]
Send data to OCS (create Stream) [3:40]
Send data to Stream and trend [6:12]
Lab summary [3:05]
[PAGE] Useful Links
Course Evaluation
How did it go?
Final Exam
Final Exam

About this course