Scenarios for Use
Use Cases for Property Transforms
Property transforms can be used in a number of ways. The following scenarios represent some typical uses for these services:
• Automate Simple Data Transformations – In some scenarios, a solution developer needs to perform a simple calculation on streaming data values. A property transform can automate the process by creating a new property configured from an existing source property. For example, a sensor may be recording raw data about the electrical current being pulled by a machine many time per second. However, a machine operator might only need to monitor the maximum electrical current every minute, or the average electrical current during a 30 second window. A property transform can be configured to calculate the data in the required form.
• Prepare Predictive Analytics Data – In this type of scenario, a solution developer plans to use data from a number of connected assets to predict the estimated time to failure of the monitored equipment. As part of creating the predictive model, the solution developer identified a number of transformed features that are required inputs for making the prediction about estimated time to failure. Property transforms can automate some of the data preparation required to transform the raw data into the required forms.
• Create Statistical Monitoring Alerts – Statistical control rules can be used to monitor whether a process or asset is operating under control. These rules enable monitoring for patterns in the data that deviate from normal or expected behaviors. Alerts can be created to notify key stakeholders of abnormal conditions that require investigation. Property transforms can be created to enable the statistical monitoring of processes or assets. How? Configure one of the Property Transform count services to monitor data values coming into ThingWorx according to rules specified in the property definition. The Property Transform count services include the following:
◦ Threshold count – Counts the number of values above or below a specified limit.
◦ Range count – Counts the number of values within or outside of a specified range.
◦ Trend count – Counts the number of values trending up, down, or in alternating directions.
In ThingWorx, configure an alert to react when the values of a property transform are outside of an established limit.
Descriptive Analytics Services or Property Transforms?
Both Descriptive Analytics Services and Property Transforms provide a way to implement common statistical calculations on property data. However, these tools are designed to handle different scenarios and use different computing resources. While Descriptive Analytics provides on-demand services in a batch scenario, Property Transform services derive value from streaming data as it enters ThingWorx. That said, in some scenarios, it can be difficult to decide which approach is more appropriate.
For an in depth comparison of the two tools, see the PTC Community article entitled
Descriptive Services or Property Transforms, which approach to use? The article examines the similarities and differences between the two approaches and provides guidelines for choosing the right approach for a given scenario.