Cumulocity IoT Analytics Builder

Analyze live streamed data from machines and devices without writing code. With Cumulocity IoT Analytics Builder, operational technicians, factory-floor engineers and analysts can build analytics on their own to improve operational efficiencies faster.

Recognizing and acting on patterns in live data from machines and devices enables you to respond to changing events while you can influence the outcome. Until now, that’s been complicated, requiring coding and deep analytics skills to create logic based on many versions of“if this, then that” calculations. Cumulocity IoT Analytics Builder makes it simple. Anyone can define self-service analytics using easy-to-connect building blocks. The result? Your team can monitor operations, receive alerts for predicted behavior, and detect problematic scenarios before situations get worse, turning problems into opportunities for rapid improvement.

Drag-and-drop to build models

No coding is required. You’ll have an intuitive interface to design models that look for matching patterns in live data coming from your machines and to take appropriate action. Simply “drag and drop” to define how to act in real time on what’s happening in your production lines or on your factory floor.

Analytics Builder models react in real time, based on your live data. These models can generate business-critical alerts, remotely control machines, validate data, pre/postprocess data, derive new measurements, or aggregate multiple measurements.

Analyze your Cumulocity IoT data in real time

Analytics Builder runs within Cumulocity IoT, working on incoming measurements, such as vibrations, humidity, pressure and temperature, as well as working with Cumulocity IoT entities like events, operations and alarms.

For example, if multiple sensors show the temperature and vibration are both increasing beyond allowable operational parameters, then stop production. If the temperature is rising, but within maximum allowable operational parameters, then raise an alarm and book a service call for the next scheduled maintenance period.

Pre-built analytics make building models quick and easy

Select from a wide-ranging library of pre-built analytics that you can piece together as blocks within models to:

• Identify threshold breaches
• Calculate averages and standard deviations
• Calculate weighted linear regression gradients
• Discover missing data
• Create custom analytics blocks using the Analytics Block SDK

Deploy models in one click

The simulator provided with Analytics Builder makes it easy to test models on historical data before deploying them into production. When a model is fit for production, simply deploy it with a single click.

Real-world use-case monitoring pressure

A manufacturer of painting robots uses Cumulocity IoT Analytics Builder to create a model that raises an alarm if there is a significant deviation between the actual paint pressure and the target paint pressure. Such an alarm could indicate too little paint has caused a gap in the painting or too much paint has caused drips.

In this example, the model continually calculates the difference between the actual pressure and the target pressure. The model averages this differential over a time window that gets reset when the robot moves on to paint the next object. The average differential is output to Cumulocity IoT as a new derived measurement and can be viewed in dashboards. If the differential exceeds tolerated limits, a Cumulocity IoT creates an immediate alarm.