The integration and usage of uncertain sensor data in workflows is a difficult problem. In this paper we describe these difficulties which result from the combination of very distinct areas. On the one hand, applications from area of measurement engineering manage sensors that capture data and annotate the data with technical meta data. On the other hand, context-aware workflows from the BPM area place high level requirements for the quality of context data that is derived from the sensor data. Between those two areas exists a gap that has to be closed by a context management and mediation system, supporting the handling of Quality of Context (QoC). To achieve this the paper presents an QoC aware architecture based on an extension of the existing Nexus Platform and a first approach for matching the workflow requirements with the sensor annotations.