Multiple-Level Predictive Control of Mobile Cyber Physical Systems with Correlated Context

Abstract:

With the increasing popularity of mobile computing, cyber physical systems are merging into major mobile systems of our society, such as public transportation, supply chain systems, health and wellness, and taxi networks. Mobile CPSs interact with phenomena of interest at different locations and environments, and where the context information (e.g., network availability and connectivity) about these physical locations might not be available. This unique feature calls for new solutions to seamlessly integrate mobile computing with physical world processes by sharing information among networked mobile CPSs. Within this context we have developed both fundamental solutions and application based solutions. Regarding CPS fundamentals, we have addressed systems-of-systems issues, human-in-the-loop control, and model predictive control. One highlight is that we have created an approach to detect and resolve sensing and actuation conflicts among independently developed CPS applications running on a generic sensing and actuation infrastructure. This work was a finalist for best paper award at ICCPS 2014. We are also developing control solutions where the human is a sensor, the system, the controller, or all of these at the same time. In the application area, we have developed multiple solutions for modeling and control of transportation systems. Results in this area include or are currently addressing: (i) reducing time for drivers to find customers; (ii) reducing time for passengers to wait, (iii) avoiding and preventing traffic congestion; (iv) reducing gas consumption and operating cost; (v) improving driver and vehicle safety, and (vi) enabling passengers to share taxis. To achieve these goals, rich contextual information (GPS readings, taxi occupancy, wireless link quality, and a priori knowledge of the normal traffic patterns per day and time of day) are collected and analyzed. Various dispatch control commands are delivered to taxi drivers. We are also investigating improving reacting to medical alarms in intensive care units. This includes classifying alarms, prioritizing them, and helping nurses if distractions affect alarm processing.