Data Processing
There are several ways of obtaining vital parameters from sensor data directly derived from the body. First, however, the raw data must be filtered and pre-processed. This project phase will examine whether the primary data should be processed directly at the sensor (edge computing) or in a server in order to efficiently conduct an AI-supported evaluation of heterogeneous sensor data.
The objective is to develop and test a data processing procedure (possibly supported by special hardware) to determine vital parameters. In addition to AI methods, concepts of self-monitoring and redundancy should guarantee reliability, even if individual sensors fail. With regard to future applications, the SanTrain project could be used for studying ways of measuring external respiratory parameters and, at the same time, blood oxygen saturation. This demonstrator would allow for the accurate and reliable measurement of signals from the skin surface, for example without a constricting chest strap. It would, for example, allow us to develop and demonstrate all modules that are necessary for data processing and for identifying vital parameters.