Real-time data processing and visualization of a hydrocarbon sensor network for hydrocarbon environmental monitoring

The development of a graphical user interface for a hydrocarbon sensor network system is presented in this paper. The software introduces two mathematical algorithms for realtime data processing and visualization: The Kalman filter is employed to remove spikes from acquired data; PCA (Principal Component Analysis) is used to reduce the dimension of the multiple sensor dataset and project the dominating differences among measured water samples on a 2D scores plot. The software also offers easy interface for simultaneous monitoring of system health status and logging activities and observations. This design greatly facilitates fast identification of anomalies, detection of sensor failure and timely collection of representative samples. Performance of the software interface is evaluated by applying real-world data collected during the Gulf of Mexico survey.