Agent-Based Simulation of Smart Beds With Internet-of-Things for Exploring Big Data Analytics

Internet-of-Things (IoT) can allow healthcare professionals to remotely monitor patients by analyzing the sensors outputs with big data analytics. Sleeping conditions are one of the most influential factors on health. However, the literature lacks of the appropriate simulation tools to widely support the research on the recognition of sleeping postures. This paper proposes an agent-based simulation framework to simulate sleeper movements on a simulated smart bed with load sensors. This framework allows one to define sleeping posture recognition algorithms and compare their outcomes with the poses adopted by the sleeper. This novel presented ABS-BedIoT simulator allows users to graphically explore the results with starplots, evolution charts, and final visual representations of the states of the bed sensors. This simulator can also generate logs text files with big data for applying offline big data techniques on them. The source code of ABS-BedIoT and some examples of logs are freely available from a public research repository. The current approach is illustrated with an algorithm that properly recognized the simulated sleeping postures with an average accuracy of 98%. This accuracy is higher than the one reported by an existing alternative work in this area.