Managing Information and Storage in Networked Cyber-Physical Systems

The integration of computer-centric control and physical systems is becoming more entrenched, as we see these cyber-physical systems (CPS) appearing in a wide variety of systems including medical systems, industrial monitoring, agriculture, avionics, amongst others. However, we see the following limitations to the wide-spread adoption of CPSs, namely inadequate information and data capabilities, distributed control of networked CPSs, low power management in mobile/wearable devices, and high performance computation capabilities for complex data manipulation e.g. real-time image processing. Research in the latter two areas is active due to the common needs in general computing particularly mobile devices. While control systems for a free-standing CPS are well understood and studied, systems to control distributed systems have not been sufficiently explored. Also, information management in a distributed CPS has not been addressed adequately in current research efforts and needs to be investigated more fully. In the remainder of this position paper, we will focus on architectures and challenges of an information management system for a CPS. As we develop CPSs with more advanced intelligence, it is clear that the capabilities of these systems must include persistent (or perceived persistent) access to information and data. This information could be data acquired by observing the environment and surroundings, information learned through inference, or static and dynamic data that is retrieved from sources external to the system. As information needs grow, it becomes critical that we efficiently manage the information retrieval and storage capabilities particularly in CPSs that are networked.