ThingWorx Analytics Training: Module 4 Part 2
This video concludes Module 4: Data Transformation & Feature Engineering of the ThingWorx Analytics Training videos. It covers Descriptive Services and Derived Properties, and how they can be leveraged to create helpful alerts and make data transformation easier.
This video does not contain any hands-on exercises, and you don't need access to a running ThingWorx Analytics environment.
This video has chapters — use the chapter menu in the player to jump to any section.
Chapter Summaries
0:01 – Descriptive Services Intro: ThingWorx Analytics 8.3 Feature Engineering
Introduces descriptive services, added in ThingWorx Analytics 8.3, which engineer meaningful features from time-series data without custom code, powered by the Descriptive Analytics microserver that must be running to enable both new microservers.
0:33 – Statistical Calculation Microserver: IoT Time-Series Aggregation
Covers the Statistical Calculation microserver and its library of services that aggregate and transform raw streaming data from IoT machine sensors, saving time versus writing custom code for calculations like mean values.
1:04 – Setup: Statistical Calculation Thing Shape & Logged Properties
Walks through attaching the Statistical Calculation Thing Shape to a thing or thing template, adding logged properties to store returned values, and feeding timed values from the Query Timed Values for Property service to compute binned distributions, confidence intervals, and five-number values.
1:58 – Statistical Monitoring Microserver: SPC Rules & Control Limits
Explains the Statistical Monitoring microserver and Statistical Process Control (SPC) rules that flag when readings breach upper or lower control limits, using services like Get Number of Consecutive Points Based on a Range against value-stream data to trigger abnormal-condition alerts.
3:08 – Chaining Service Calls: Query Timed Values & Five-Number Summary
Demonstrates chaining two service calls in code, setting timedValues from Query Timed Values for Property with a maxItems of 500, then assigning the mandatory result variable from Calculate Five-Number Property Values to return minimum, first quartile, median, third quartile, and maximum.
4:22 – Trend Detection: Consecutive Increasing/Decreasing/Alternating Points
Shows the Get Number of Consecutive Points Following a Trend service returning the largest run of increasing, decreasing, or alternating points via a trend parameter, illustrated with plotted examples returning five decreasing, six increasing, and alternating point counts.
5:45 – Derived Properties: Codeless Transforms in ThingWorx Analytics 8.4+
Introduces derived properties, available in ThingWorx Analytics 8.4 and higher, a codeless way to transform raw data by adding a Transformed Property with a Logged checkbox, a source property, a calculation type (minimum, count, average), and data collection parameters.
6:36 – SPC Use Cases: Torque Monitoring & Defect Detection Transforms
Presents statistical monitoring transform use cases, including tracking maximum torque during a pressing operation and applying SPC rules to automated quality inspection defect counts using expected values and standard deviation for early trend and abnormality detection.
8:00 – Descriptive Services vs Derived Properties: Batch vs Streaming
Compares the two approaches: descriptive services run through the Descriptive Analytics microservice on batched data via Query Timed Values for Property, while derived properties are created codelessly through the Properties screen or API and applied to streaming data, though extra transforms exist only in descriptive services.
8:42 – Wrap-Up: Upcoming Modules & PTC Customer Success
Closes the ThingWorx Analytics training module with a preview of upcoming modules and an invitation to contact a PTC Customer Success representative for more information on using ThingWorx Analytics.