Embedded control is a key product technology differentiator for many of the Dutch high-tech industries. The strong increase in complexity of embedded control systems combined with the occurrence of late changes in control requirements, results in many timing performance problems showing up only during the integration phase. This results in extremely costly design iterations, severely threatening the time-to-market and time-to-quality constraints.

In the Carm 2G project this integration problem is attacked systematically through the (largely automated) construction of executable models (see picture below). The key approach is to separate the logic of the embedded control application from the execution platform on which it is deployed. The resulting models yield a high-level overview and provide system-wide insight in timing bottlenecks. They further allow rapid exploration of alternatives for optimization of timing performance (by adapting the application, the execution platform or the mapping).

The Carm 2G project has demonstrated the effectiveness of the performance prediction and optimization method by applying it to a complex performance-critical subsystem of a wafer scanner. The application of the method has resulted in more than a dozen improvement proposals yielding a huge overall timing performance gain and resulted in a development roadmap of the execution platform.