Orchestrating Task Chains to meet End-to-End Delays in Complex Embedded Systems

Embedded systems are at the heart of the ongoing digitalization that transforms all areas of society. Consequently, their complexity is steadily rising.

Many of these systems are safety-critical by nature and subject to various timing constraints that need to be met at runtime, requiring formal evidence of correctness.

The most crucial timing constraint for future applications is the end-to-end latency experienced by data passing through a chain of semantically related tasks.

Current approaches verify such timing constraints at the final development stages, which can lead to costly design iterations and is intractable for future applications, where more than a thousand task chains must be orchestrated over complex computing platforms.This project radically rethinks how end-to-end latencies are treated during the design of safety-critical embedded systems.

This is done in three steps.

Formal timing analysis is elevated to be applicable in the early design phases of heterogeneous software and hardware systems. Necessary constraints for the application are identified that guarantee meeting the task chains´ timing constraints. And finally, the identified constraints are integrated at the system level.

With its theoretical and practical aspects, the project delivers a crucial contribution required to design future embedded systems for a more safe and more sustainable future. The strong background of the applicant makes him uniquely qualified to address this research challenge.