Ensembles of Human and Digital Sensors: Faithful Representations Through Hybrid Sensing Systems

Zeiss Roman, Hovestadt Christian, Chasin Friedrich

Abstract

Digital sensors capture a growing range of real-world phenomena with increasing robustness and precision. However, we observe that the overreliance on digital sensors comes with technical and economic feasibility constraints limiting which aspects of real-world phenomena information systems can represent. Using representation theory as a theoretical lens, we argue that real-world phenomena can be represented more faithfully in IS if digital sensors are complemented by human sensors working in concert as a hybrid sensing system. We use a mixed-method research approach to study the design, operation, and impact of hybrid sensing systems at a German bike sharing provider that relies on such a system to monitor the location and availability of its fleet. Through this research, we aim to bring the origin of data into focus of information systems design and present a sociotechnical solution to the fundamental problem of limited observability of user misbehavior in sharing economy business models.

Keywords
Digital sensors; human sensors; sensing systems; representation theory; internet of things; data sources; data generation