1. Time-Aware Concepts for edge computing: Primarily, we will develop a better time-aware support for applications. Thus, we need to implement a novel time-aware semantics for all analytics components enabling deterministic and therefore traceable execution of analytics workflows independent of the processing in the sensor devices
  2. Predictive analytics plays a major role in many predictive maintenance user scenarios. It is vital for these scenarios to predict a potential failure “early enough” to avoid the event of a critical failure. In this way, we reduce the failure’s impact to a minimum.
  3. Overall, the distributed decision making will allow users to make decisions faster and more accurate by providing meaningful knowledge models and coherent decisions for humans.

The next generation of sensor networks’ analytics methods for many industries will improve all steps within their business processes to be fast, versatile and competitive. To make the right decisions on time, it is vital to move the data processing away from centralized nodes into the edge of networks, aiming at reducing the latency and minimizing essential information exchange between sensors and the analytics components.

The main objective is to lower the integration barrier of time-aware data-analytics in the next generation of industrial application by researching on predictive analytics and cognitive methods using time-aware concepts. This will make the establishment of data driven applications much easier and cheaper.