SBIR Phase II: Bridging Fugitive Emission Monitoring Gaps via Artificial Intelligence and Hybrid Sensors

The broader/commercial impact of this Small Business Innovation Research (SBIR) Phase II project would be in empowering oil & gas producers to cost-effectively detect emissions from their facilities early through sub-weekly, high-resolution AI-powered satellite scans integrated with continuous 24/7 onsite monitoring. This hybrid approach delivers comprehensive leak detection, facilitates swift mitigation, and generates credible certification data enabling premium pricing for produced natural gas. The primary, high-risk technical innovation of this work is a deep learning model designed to automatically detect and quantify asset-level specific emissions in freely available, 30-m-resolution satellite imagery not originally designed for detecting those emissions (Landsat, Sentinel-2). The core objective is to integrate training imagery coincident with known emissions below the theoretical detection threshold of these satellites when using traditional methods (200-500 kg/hr under ideal conditions). With proprietary, high-quality training data, the model may achieve detection thresholds similar to commercial satellites (<100-150 kg/hr), but at a fraction of the cost and at much higher temporal frequency. This model will be used to monitor an operator’s complete suite of onshore assets at sub-weekly intervals, which will be integrated with continuous ground-sensor observations at a subset of sites. The second innovation is a machine learning model to predict daily emission rate probabilities at each asset – based on asset-level details, ground sensors, and satellite-derived methane concentrations – for targeted monitoring and mitigation. The scope of this Phase II project encompasses the development, validation, and field-testing (controlled-release study) of a holistic platform for monitoring. The core intellectual contribution lies in bridging the spatial and temporal gaps in current fugitive gas monitoring paradigms. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.