Leveraging Distributed Intelligence for Low-Latency Decision-Making in Cyber-Physical Systems

Abstract

Cyber-physical systems include three stages where data is converted into information, which in turn, becomes knowledge. The conversion from data to information occurs at constrained smart devices, while the conversion from information into knowledge happens at specialized artificial intelligence (AI) applications on the cloud. In mist computing scenarios, with the goal of reducing the latency associated with knowledge extraction, the devices play the role of converting information into knowledge. Specifically, a device collects aggregated information captured at other devices and uses it for actuation. One critical goal is to maximize the lifetime of the network to make sure there are no devices that consume proportionally more power than the rest. To this end, this paper introduces a “distributed intelligence” mist computing mechanism that recycles well-known wireless sensor network (WSN) routing technologies to maximize the overall network lifetime. In addition, this algorithm pairs with standard internet of things (IoT) networking protocols to minimize the latency of the decision-making process.