Jyotsna Sharma

Marine oil spills necessitate prompt identification to minimize environmental and economic repercussions. Although satellite-borne Synthetic Aperture Radar (SAR) affords crucial all-weather monitoring, extracting insights from these data remains complex. Conventional deep neural networks typically require large training corpora and suffer from high latency. To circumvent these limitations, a pixel-level quantum-enhanced Support Vector Machine (SVM) bagging ensemble is proposed. The architecture uses quantum annealing to determine the optimal support vectors for independent weak SVM classifiers trained on constrained data subsets, prior to their classical aggregation. This methodology is evaluated on Sentinel-1 imagery using both simulated and physical quantum annealing hardware. The quantum-enhanced framework demonstrates efficacy comparable with a strict classical baseline, achieving an Intersection-over-Union (IoU) of 0.60 and a balanced accuracy of 0.89. Complementary experiments employing gate-based quantum circuits yielded equivalent segmentation quality, although the annealing paradigm provided accelerated inference speeds. Collectively, these findings validate the viability of quantum-assisted, low-data segmentation architectures for near-real-time environmental surveillance.

Summary: Modern managed pressure drilling (MPD)-based well control operations still face challenges with effective gas influx detection and management, including the application of appropriate kick detection thresholds across different mud systems, and the limited understanding of mass-transfer dynamics and multiphase behaviors when nonaqueous drilling fluids are involved. Through a comprehensive full-scale comparison of gas influx signatures between water-based mud (WBM) and synthetic oil-based mud (SBM) systems with matched rheological properties, the objective of this study is to characterize the mass-transfer-dependent effects on gas influx behavior. Experiments were conducted at Louisiana State University’s Petroleum Engineering Research, Training, and Testing (LSU’s PERTT) Laboratory using a full-scale wellbore equipped with industry-standard MPD systems and distributed fiber-optic sensing (DFOS) technology. A systematic rheological formulation and verification protocol was used to develop WBM and SBM systems with closely comparable flow-governing rheological behavior over the wellbore-relevant shear-rate range. This rheology-controlled design reduces hydrodynamic confounding and enables a controlled assessment of mass-transfer-dependent effects on gas influx behavior. Controlled gas influxes were introduced under various operating conditions, with continuous monitoring of pressure, temperature, and flow parameters throughout the wellbore. Key multiphase phenomena, including gas slippage, dispersion behavior, and mass transfer between phases, were measured and analyzed. The experimental results revealed significant differences in gas influx signatures that challenge the industry practice of applying uniform kick detection thresholds across different mud systems. DFOS measurements provided spatial and temporal resolution of gas influx signatures, showing that SBM systems, despite having closely comparable rheological parameters, exhibited different gas dissolution rates and bubble dynamics compared with WBM. The high-resolution data collected throughout the wellbore showed that conventional models underestimate the complexity of gas-SBM interactions, particularly regarding gas dissolution, reevolution (outgassing), and their time-dependent nature. An advanced transient multiphase flow simulator was used to compare predictions with the experimental observations, enabling refinement of mass transfer submodels and slip correlations in a decoupled manner for each mud system. This improved modeling capability supports the development of enhanced early warning algorithms that account for the unique mass transfer dynamics. This research presents the first comprehensive full-scale experimental comparison between rheology-controlled WBM and SBM systems, providing a controlled full-scale data set that highlights the mass-transfer-dependent effects on gas influx behavior. The unique data set from the full-scale facility with advanced MPD and sensing capabilities addresses a knowledge gap in multiphase flow behavior in different drilling fluids, advancing both theoretical understanding and practical well control applications.

Cesium lead bromide (CsPbBr₃, CPB) is a promising all-inorganic perovskite for radiation detection; however, CsPbBr₃ single crystals exhibit unstable electrical behavior under low-dose irradiation. This study advances CsPbBr₃-based detection by evaluating two sensing architectures integrated with supervised machine learning (ML): a direct electrical single-crystal detector and a strain-based CPB-polymethyl methacrylate (CPB-PMMA) coated Fiber Bragg Grating (FBG) sensor. Gamma radiation experiments were performed using a low-dose Cs-137 and a high-dose Co-57. The high-resolution signals were analyzed using an ensemble-based algorithm enhanced with temporal signal components and derivative features engineering to classify radiation states and identify radiation sources. For the single-crystal detector, a bagging tree classifier achieved 85.1% accuracy for Cs-137 and a superior 95.8% accuracy for Co-57. Subspace discriminant ensemble results further demonstrated asymmetric transferability, with models trained on Cs-137 generalizing strongly to Co-57 (97.2% accuracy), while the reverse transfer performed more modestly (82.7% accuracy). Additionally, a multi-class RUSBoosted ensemble distinguished between background and isotopes with 85.5% accuracy. In contrast, the CPB–PMMA FBGs sensor effectively improved in low-dose detectability through a bagging tree model, achieving 96.0% accuracy for Cs-137. The results demonstrated that strain-induced photonic transduction significantly compensates for the limitation of direct electrical readouts in low-dose regions. Collectively, these results show that single crystal CsPbBr₃ is optimal for high-dose environments, while FBG-based photonic coupling extends operational capabilities to low-dose regimes with appropriate ML data processing support, highlighting CsPbBr₃’s potential as a versatile material suitable for broader and more extreme environments.

The integration of quantum science with optical fiber technology is redefining the limits of precision measurement and distributed sensing. This review provides a comprehensive overview of recent advances in quantum‐enhanced fiber‐optic sensing, focusing on how quantum light sources, quantum materials, and quantum algorithms are converging to improve sensitivity, selectivity, and dynamic range in distributed fiber‐optic sensing (DFOS). Developments in entangled and squeezed light as well as quantum‐defect and two‐dimensional materials are shown to enable measurements of temperature, strain, pressure, magnetic field, and chemical composition beyond the classical shot‐noise limit. The exploration of quantum and quantum‐inspired algorithms to improve sensing apparatuses and frameworks and help alleviate massive data bottlenecks is highlighted. Insights drawn from magnetometry, light detection and ranging (LiDAR), optomechanics, photonic computing, tensor networks, and quantum‐enhanced imaging illustrate how cross‐domain quantum technologies can inform next‐generation fiber sensing architectures. The review also identifies key challenges, such as efficient source–fiber coupling, coherence preservation over long distances, and scalable hybrid integration and outlines future pathways toward robust, field‐deployable quantum‐enhanced DFOS for environmental monitoring, biomedical diagnostics, and infrastructure security.

Summary: In an effort to enhance safety and efficiency in offshore drilling operations, we address critical gaps in the current understanding and application of leading indicators for loss of well control (LOWC) events. Existing indicators often fall short in predicting LOWC events accurately and in a timely manner due to a lack of integration of real-time data, which is crucial for proactive risk management. Our study aims to bridge these gaps by identifying robust leading indicators and incorporating them into an advanced real-time detection and response system. The research began with a comprehensive analysis of existing LOWC events to identify the key leading indicators. Using advanced data analytics techniques, including data assimilation (DA) and change-point Bayesian networks (CP-BN), we have developed an integrated real-time kick detection and state estimation framework and evaluated the effectiveness and reliability of the identified indicators. This evaluation was supported by extensive testing, including real-time and offline simulations, as well as full-scale experiments where operational parameters were adjusted to replicate various drilling scenarios. Several key leading indicators were identified based on this study, such as abnormal pressure fluctuations, unexpected changes in drilling fluid properties, and precursors detected through distributed fiber-optic sensing (DFOS) data. We conducted a series of surveys and engaged with industry professionals to gather practical insights and feedback to further validate these findings and enhance the applicability of our results. The integration of these indicators into the kick detection and state estimation framework significantly enhanced early detection capabilities. Experimental testing and assessment of the integrated framework in a full-scale system have demonstrated its operational viability, markedly improved preventive safety measures, and contributed to the reduction of overall system risks. This research represents an advancement in operational safety for offshore drilling operations, facilitating the development of proactive approaches that enable drilling crews to respond more effectively to emerging threats and reduce the likelihood of LOWC incidents. It serves as a foundational study for continued safety enhancements in the oil and gas industry, encouraging the ongoing refinement of safety practices.

This study presents the development and characterization of CsPbBr3 (CPB) and CPB-poly­(methyl methacrylate) (CPB-PMMA) composite coatings on Fiber Bragg Grating (FBGs)-based sensors for high-sensitivity gamma radiation sensing. CPB precipitates were synthesized using a solvent-based method and uniformly deposited onto the FBGs. The incorporation of PMMA into the CPB matrix enhanced both mechanical stability and adhesion to the FBG surface. Spectral analysis revealed significant Bragg wavelength shifts in response to gamma radiation exposures, indicating strain-optic variations induced by radiation–matter interactions. Comparative investigations between uncoated, CPB-coated, and CPB-PMMA-coated FBGs confirmed that the coatings significantly enhance strain sensitivity and stability. The incorporation of PMMA modified the mechanical response, influencing residual stress and strain attenuation. The advantage of using fiber optic sensors includes the ability to enable multiplexed sensing across large areas, immunity to electromagnetic interference (EMI), and operability in high-temperature environments. Additionally, CPB and CPB-PMMA coatings demonstrated enhanced sensitivity under UV exposure, further highlighting their potential for advanced optical sensing applications in both radiation and UV-intensive environments. These results demonstrate the potential of CPB-based coatings for radiation monitoring in extreme environments, including in nuclear facilities, space missions, and near medical instruments using radiation sources. The findings provide a foundation for further optimization of perovskite-polymer composites to enhance sensor performance and long-term durability in radiation-intensive applications.

CO₂ sensing is vital for environmental monitoring, industrial applications, and climate change research. This study presents an innovative numerical framework, validated through experimental results, to assess the performance of fiber-optic CO₂ sensors coated with Metal-Organic Frameworks (MOFs). An advanced numerical model has been developed that integrates Maxwell Garnett effective medium theory, Tauc Lorentz model and N-layer matrix method to evaluate the performance parameters of MOF porous structure and impacts of gas interactions. The predicted numerical results were compared with published experimental data for two different MOFs, showing a transmittance discrepancy of 3% for ZIF-8 and 1% for Co/ZIF-8, with sensitivity discrepancies of around 1% for both MOFs. The strong agreement between the calculated and experimental values for transmittance and sensitivity provides validation to this advanced numerical model. This prediction of experimental outcomes with this advanced numerical model adds novelty to this work. Furthermore, the refractive index and extinction coefficient derived for these MOFs closely align with experimental trends, reinforcing the model's validity. Additionally, the study investigates the influence of key parameters such as the porosity of the MOF structure, the numerical aperture and core diameter of the optical fiber, and the coating length of the MOF layer on performance metrics like optical transmittance and full width half maxima (FWHM). This comprehensive framework provides a valuable tool for optimizing sensor design, enhancing the development of MOF-coated CO₂ sensors, and offering guidance for future experimental studies and field implementations.