You will need ThingWorx Foundation installed and running before installing ThingWorx Analytics Server or ThingWorx Platform Analytics.

  • ThingWorx Analytics Server or ThingWorx Platform Analytics may be installed independently of each other.

ThingWorx Analytics Server:

ThingWorx Analytics Server includes everything needed to automatically create machine learning models as well as the infrastructure to use models at scale across multiple processors. This download includes two extensions that, when imported into a ThingWorx Foundation server, add an easy to use web interface for creating and deploying machine learning models.

ThingWorx Platform Analytics:

Two components are included with the ThingWorx Platform Analytics download:

  • Descriptive Analytics provides common statistical calculations to ThingWorx Foundation including: Maximum, Mean, Median, Minimum, Mode, Standard Deviation, Threshold Count, Range Count, and Trend Count.
  • Property Transform microserver provides the ability to perform calculations in real time on streaming data before it is evaluated by ThingWorx Foundation.
    • These calculations are performed outside of the ThingWorx Foundation server and require both an Apache Flink cluster and a RabbitMQ message exchange.

Supported Operating Systems (64-bit only):

  • Red Hat Enterprise Linux 8.2
  • Ubuntu 18.04
  • Ubuntu 20.04
  • Windows Server 2016
  • Windows Server 2019

Supported Browsers:

  • Microsoft Edge Chromium 89
  • Google Chrome 89
  • Firefox ESR 78
  • Safari 14

Key Benefits

Automate Analytical Processes

No need to be experienced in statistical analysis or complex mathematics. The AI and Machine Learning technologies used in ThingWorx Analytics automate much of the complex analytical processes involved in creating data-driven insights for your IIoT application.

Digital Simulations

Simulate behavior of physical products in the digital world and incorporate simulation models into your solution. Utilize the knowledge of these models while the product is operating in the real world.

Predictive Modeling and Scoring

Using predictive analytic algorithms and machine learning techniques, ThingWorx Analytics analyzes data from connected devices to find patterns in the data and generate a prediction model. The prediction model in your IoT solution makes predictions about expected future outcomes so you can optimize your business processes.

Real-time Anomaly Detection

Anomaly detection allows you to observe data from a device, learn about the typical state, and monitor for data points that fall outside of the expected range. Anomaly detection outputs can easily be integrated into an IIoT solution to trigger alerts and help your application users identify when to take action.

Descriptive Analytics

Descriptive Analytics provides services that perform common statistical calculations and facilitate statistical monitoring versus historical data. The services provided include operations such as Maximum, Mean, Median, Minimum, Mode, Standard Deviation, Threshold Count, Range Count, and Trend Count.

Transform Microservices

The Property Transform microserver provides on-demand transformation services that can be used to derive value from streaming data entering ThingWorx, as opposed to Descriptive Analytics' historical limitation.


  • Automates complex calculations and generates predictions, simulations, and prescriptions.
  • Utilizing data from your Things as well as historical data as a learning source, Analytics Server uses machine learning to build and validate predictive models without assistance from a statistician.
  • Employs sophisticated predictive modeling algorithms in order to determine the best algorithm to use for each data set and predictive topic. These predictive models are instantly usable by ThingPredictor and ThingOptimizer to perform predictions, simulations and determine recommendations.
  • Dramatically reduces or eliminates the need for an expert team in modeling algorithms or technologies.
  • ThingWatcher is a Java API that can be used to build anomaly detection functionality into your IoT applications.
  • Capturing data from an edge device, ThingWatcher automatically observes and learns the normal state pattern and uses those results to train and generate a model for detecting anomalies without the need for setting rules or applying pre-calculations.
  • Once a model is trained and validated, ThingWatcher scores new streaming data from the monitored device against the model to detect possible anomalies in real-time.
  • ThingWatcher can be used at any location, cloud or edge. By deploying ThingWatcher at the edge, you can monitor a stream of high-speed data that would be difficult or impossible to process in the cloud.
  • ThingPredictor provides the predictive scoring capabilities of ThingWorx Analytics.
  • It uses prediction models generated by ThingWorx Analytics Server or equivalent Predictive Model Markup Language (PMML) compliant prediction model generation tools to examine a dataset and predict results for each record based on similarities to records analyzed during model training.
  • Using ThingPredictor, you can subscribe Things to relevant outcome-based predictions (time to failure, errors per hour, etc.) and displays results in context through any ThingWorx-powered solution or experience.
  • ThingWorx Analytics predictive scoring can be accessed through a REST API Service, Analytics Builder, or Analytics Manager.
  • ThingOptimizer provides the prescriptive scoring and optimization capabilities of ThingWorx Analytics and allows you to expand your analytical processes beyond predicting outcomes to seeing how modifications might affect results by automatically identifying influencing factors.
  • Perform ad-hoc outcome simulations before initiating an action and identify optimal settings to maximize results or minimize risks.
  • Identify key contributing factors associated with predicted outcomes.
  • ThingOptimizer uses prediction models generated by ThingWorx Analytics Server or equivalent PMML-compliant prediction model generation tool.
  • Provides a user interface for simple, intuitive interaction with data.
  • Converts complex data readings into simple-to-understand information and simplifies the advanced analytics process, helping interpret information faster.
  • Supports, data and metadata uploading, predictive model creation (training and scoring), visualization of data analytics (profiles, signals), as well as data filtering.
  • Allows you to deploy and execute computational models from external applications within the ThingWorx platform.
  • Leverages product-based analysis models developed using PTC and third-party tools while building solutions on the ThingWorx platform.