MARS: A Flexible Real-Time Streaming Platform for Testing Automation Systems

Raphael Eidenbenza, Alexandru Mogab, Thanikesavan Sivanthic and Carsten Franked
ABB Corporate Research, Baden-Dättwil, Switzerland.


Trends in industrial automation systems are placing more importance on using streams of digitized data to perform various automation functions in real-time, e.g., power, process, and factory automation. To ensure high reliability and availability, individual devices or (sub-)systems thereof need to be tested with respect to their expected real-time behavior in the system context at various stages during product and project life cycle. In this paper, we introduce a real-time streaming platform called MARS that provides the means to monitor (M) high frequency data streams, analyze (A) them with respect to their properties, record data traffic (R) based on user-specified rules, and simulate data traffic (S) and in general the behavior of certain functions of one or more devices according to highly customizable scenarios. We present the design principles of MARS, the real-time software architecture, and evaluation results.

Full Text (PDF)