• AUTOMATION 21XX
Navigation : EXPO21XX News » Automation & Robotics » SICK Prediction Box enables highly available operations by using intelligent prediction models

SICK Prediction Box enables highly available operations by using intelligent prediction models

Predicting services: The Prediction Box presents information about future service events in a simple and understandable way on a dashboard or mobile.
Photo by Sick AG

Making even better use of data for service – that is the task of the SICK Prediction Box. Intelligent predictive models generate valuable information that creates high availability and prevents unplanned operational statuses, thus ensuring cost savings. Machine utilization and service calls can be planned even more effectively.

When it comes to maintenance, foresight is currently the best strategy for avoiding downtime and increasing availability while improving product and process quality. Where reactive or cyclical service calls have previously been necessary, SICK is now launching a solution with the Prediction Box that makes demand-oriented service possible by means of intelligent prediction models.

“The Prediction Box makes it possible to predict device data flows of and thereby increase productivity with early action. What makes it special is that the relevant data is processed using various mathematical models and combined with device and application knowledge to calculate predictions. By working intelligently with the data, we are able to create new information that offers very special added value,” explains Max Dietrich, Product and Application Manager of Smart Data Solutions. Based on these intelligent predictive models, users know when incidents are most likely to occur, allowing them to plan the optimal timing and scope for service.

Data direct from the sensor into the box

The Prediction Box obtains the data directly from the (SICK) sensor system. The customer then only has to connect the sensor to the cloud service (smartservice.sick.com) using the Smart Service Gateway (TDC-E). Condition Monitoring is not a mandatory requirement for this, but rather a useful addition. The Prediction Box records the data relevant for predicting future events, intelligently calculates the values, and presents the newly generated information in a simple and understandable way on a dashboard or mobile. Users can access the data from anywhere in the world. The newly created data can also be included in third-party systems via a defined interface. To ensure high prediction quality, prediction models are trained and validated to match the data flow. This makes it possible to predict known events in devices, process or automation applications with high probability and accuracy.

For more information, please visit http://www.sick.com.