Cyber-Physical Systems (CPS) comprise one of the largest information-technology sectors worldwide which is a driver for innovation in other crucial industrial sectors such as health industries, industrial automation and robotics, avionics and space. Nowadays, the embedded compute infrastructure of complex CPS is based on heterogeneous multi-core or many-core systems, which are distributed, and connected via complex networks. Manufacturing companies of distributed Cyber-Physical Systems (dCPS), such as ASML, Océ, and Philips, are facing serious challenges with respect to designing their next generation lithography scanner machines, industrial printers, and interventional X-ray machines, respectively. Typically, these machines are very complex dCPS that integrate and interconnect a possibly large number of subsystems containing multiple dependent compute nodes (hardware and software components) that perform different tasks, e.g., data processing, control, monitoring, logging/reporting, etc., thereby realizing a wide range of functionality and features. Designers of such systems need quick answers to so-called “what-if” questions with respect to possible design decisions/choices and their consequences on system performance, cost, etc. This calls for efficient and scalable system- level design space exploration (DSE) methods for dCPS that integrate appropriate application workload and system architectures models, simulation and optimization techniques, as well as supporting tools to facilitate the exploration of a wide range of design decisions. However, such DSE technology for complex dCPS does currently not exist. Therefore, in this research proposal, we address the following main research question: How can we perform efficient and effective DSE for complex, distributed cyber-physical systems? More specifically, this project proposal concerns research on DSE techniques for complex dCPS in which exploration is performed in four dimensions: i) the software workload using workload models that can capture software process activities with a granularity that will allow to explore potential workload balancing techniques, degrees of parallelism, etc., ii) the platform architecture considering potential distributed architectures to which the workload can be mapped to (e.g., deployment of ‘fat’ versus ‘thin’ compute nodes, exploitation of heterogeneous and possibly domain-specific system architectures, etc.), iii) the mapping of software processes to platform resources, and iv) the exploration of different system configurations to adapt the systems to specific workloads and platform architectures. Performing such DSE should lead to system optimization in terms of objectives like increased system performance and decreased system costs.
This work is done in collaboration with Leiden University and ASML and has been funded by The Netherlands Organisation for Scientific Research (NWO).