Extracting knowledge on slope behaviour from acoustic emission measurements

Early warning systems for slope instability need to alert users of accelerating slope deformation behaviour to enable safety-critical decisions to be made. Field trials of acoustic emission (AE) monitoring of slopes have demonstrated conclusively that generated AE rates are proportional to slope deformation rates, and AE monitoring can be an effective approach to detect accelerating movements and communicate warnings to users. AE is becoming an accepted monitoring technology for geotechnical applications; however, challenges still exist to develop widely applicable interpretation strategies. In this paper, data from a field trial at Hollin Hill, North Yorkshire, UK and a large-scale experiment are used to develop strategies to extract knowledge on slope behaviour from AE measurements. Machine learning approaches for automated interpretation (warning trigger levels and quantifying rates of slope movement) are developed and demonstrated. A conceptual framework for extracting knowledge from AE measurements for use in decision-making is presented.