Toxic gas boundary area detection in large-scale petrochemical plants with industrial wireless sensor networks

Abstract

Industrial WSNs are evolving to become the key interconnection between management and factory products in large-scale petrochemical plants. Apart from improved manufacturing, asset tracking, and robotic applications, toxic gas detection is one of the major issues in petrochemical plants, since toxic gas leakage can severely threaten the safety of first-line working staff. Continuous object detection, one of the major applications in WSNs, has become an important research topic in large-scale industry. This article overviews continuous object detection techniques that have emerged in recent years. Most of the research focuses on the estimation of the toxic gas boundary. However, an accurate boundary is less likely to be detected due to the nature (e.g., invisibility, fast movement, and changing shape) of toxic gas. Thus, it is essential to ensure the boundary area rather than only the boundary of the toxic gas. We propose a novel boundary area detection technique with planarization algorithms like RNG and GG. Exhaustive simulation studies enable us to find an optimal trade-off point between the cost of a number of deployed sensor nodes and the accuracy of the estimated toxic gas boundary area size.