Project Concept

This project will introduce anomaly detection using ThingWorx and how to construct complex alert systems.

Following the steps in this guide, you will create complex alert systems, configure an anomaly detection syste, and utilze the ThingWatcher utility. ThingWatcher functionality compares two sets of data, a validation set (collected during the Calibrating phase) and a test set (data streaming from a remote device).

We will teach you how to monitor equipment critical to your business, including devices deployed in the field. With ThingWorx, you have always been able to add an alert to a property that will be triggered when the value goes above or below a defined threshold.

Project Info

Type How To
Level Intermediate
Est. Time 60 Minutes
Published June 14, 2018
Author Al Smith

You'll learn how to

    • Configure the Anomaly Detection subsystem
    • Run the Results and Training Microservices
    • Add Anomaly Alerts to numeric properties
    • Train the Anomaly Detection system
    • Use Anomaly Detection to monitor incoming time-series property values
Register to Start Project

Things used in this project

Environment:

This guide requires a connection to the ThingWorx Foundation.

It assumes that you are utilizing the 30-day hosted Foundation server or the 120-day downloaded trial.

Software components:
Previous knowledge:

MONITOR DEVICE PERFORMANCE & CREATE ALERTS

Outline
  • Step 1: Set Up and Run ThingWatcher Microservices
  • Step 2: Configure Anomaly Detection Subsystem
  • Step 3: Add Anomaly Alert to Property
  • Step 4: Train Anomaly Alerts
  • Step 5: Monitor Properties using Anomaly Alerts
  • Step 6: Use the Sample Anomaly Dashboard Mashup
  • Step 7: Troubleshooting
  • Step 8: Next Steps