DOI: 10.36347/sjet.2017.v05i12.009

Pages: 715-718

Slope failures in mines are commonly preceded by a creep deformation curve comprising primary, secondary, and tertiary phases. The final tertiary (accelerating) phase provides the strongest precursor signals of imminent collapse, but reliably identifying its onset (the Onset of Slope Failure, or OOSF) in noisy monitoring data remains challenging. Traditional early-warning systems rely on fixed displacement or velocity thresholds, but once those limits are reached, the lead-time before failure is uncertain. We propose a new algorithmic framework to detect the OOSF by analyzing real-time displacement time-series with a moving-window multi-criteria test. The method applies sequential checks on short windows of data. Only when all checks are satisfied is the window start marked as the OOSF. This systematic detection of tertiary creep onset could improve the early forecasting of slope failures.