Multi-location gas sensing is critical for a large number of civil infrastructure and environmental applications, including energy grid monitoring and the surveillance of dangerous processes such as carbon sequestration or areas such as coal mines and landfills. For such applications, which generally involve a harsh sensing environment with significant safety concerns, fiber optic gas sensing is often the only solution. This is because fiber gas sensors consume little energy and have no need for batteries or other potentially "unsafe" electrical power supplies. Traditionally, it has been difficult to combine a large number of fiber gas sensors together for multi-location sensing. This project aims to overcome this challenge by developing an adaptive gas sensor network, where the behavior of each gas sensor can be individually tuned by changing the spatial profiles of optical signals within the sensor network. This makes large scale integration of 10 to 100 gas sensors possible. As a result, this project has the potential to significantly reduce per-sensor-cost for gas sensing and impact a myriad of engineering applications that involve explosive or dangerous gases. This project also includes a comprehensive education and outreach plan targeting underrepresented groups at all levels.
Current approaches for fiber optic gas sensing rely on using single mode fibers. This project aims to overturn this conventional paradigm, and to demonstrate a first-of-its-kind adaptive mode-division-multiplexing (MDM) gas sensor network based on few mode fibers. In this approach, the sensing characteristics of individual gas sensors can be dynamically modulated by using adaptive optics (AO) to control the mode composition of interrogation signals. This sensor network relies on the highest order fiber mode for gas detection and utilizes the lowest order fiber mode for low loss signal transmission. This unique design allows one to resolve the fundamental conflict between high gas sensitivity, which requires large gas attenuation for individual sensors, and large scale multiplexing, which demands low-loss power transmission for the entire sensor network. Specifically, this project aims to experimentally demonstrate an AO-MDM sensor network that contains at least 10 gas sensors for acetylene monitoring, and theoretically explore the possibility of constructing a highly multiplexed sensor network with 100 or even 1000 gas sensors.
