RTEMP: Exploring an end-to-end, agnostic platform for multidisciplinary real-time analytics in the space physics community and beyond

Abstract:

Large-scale, real-time, sensor-driven analytics are a highly effective set of tools in many research environments; however, the barrier to entry is expensive and the learning curve is steep. These systems need to operate efficiently from end to end, with the key aspects being data transmission, acquisition, management and organization, and retrieval. When building a generic multidisciplinary platform, acquisition and data management needs to be designed with scalability and flexibility as the primary focus. Additionally, in order to leverage current sensor web technologies, the integration of common sensor data standards (ie. SensorML and SWE Services) should be supported. Perhaps most important, researchers should be able to get started and integrate the platform into their set of research tools as easily and quickly as possible.

The largest issue with current platforms is that the sensor data must be formed and described using the previously mentioned standards. As useful as these standards are for organizing data, they are cumbersome to adopt, often restrictive, and are required to be geospatially-driven.

Our solution, RTEMP (Real-time Environment Monitoring Platform), is a real-time analytics platform with over ten years and an estimated two million dollars of investment. It has been developed for our continuously expanding requirements of operating and building remote sensors and supporting equipment for space physics research. A key benefit of our approach is RTEMP’s ability to manage agnostic data. This allows data that flows through the system to be structured in any way that best addresses the needs of the sensor operators and data users, enabling extensive flexibility and streamlined development and research.

Here we begin with an overview of RTEMP and how it is structured. Additionally, we will showcase the ways that we are using RTEMP and how it is being adopted by researchers in an increasingly broad range of other research fields. We will lay out a roadmap describing our path forward in the development of RTEMP over the next year, including the anticipated incorporation of new in line analytics for carrying out feature extraction and autonomous decision making “on the fly”. Our goal, within a year, is to be operating the world’s first true sensor web (as opposed to sensor network) for geospace research.