In IoT/M2M platforms such as those defined by ETSI and oneM2M, all data collected by sensors are transmitted and stored in the platforms before they are processed. These data consume much of platform resources including network bandwidth, storage, CPU and memory. Nevertheless, in reality not all these data are needed by the IoT/M2M applications. In this paper, we propose to add a Streaming Data Preprocessor in the IoT/M2M platform to deal with fast velocity and large volume data from the IoT/M2M sensors. We experiment and compare the data processing capability of our enhanced platform with the one without a preprocessor based on an ETSI M2M-compliant platform OpenMTC. Our experiment was conducted for a factory management use case from which we show that our enhanced system demonstrates better performance.