I’m looking to round out CPD with a course that ties mine pit slope stability assessment directly to 3D geological model uncertainty — ideally something hands-on with limit-equilibrium/probabilistic methods (e.g., Slide2) and integration of structural mapping, lab shear strength, and downhole televiewer data. Has anyone taken a program in the last 6–12 months that teaches data workflow from core logging to factor-of-safety reliability, not just theory?
I did the Rocscience Slide2/Slide3 probabilistic short course in May; it tied pit slope LEM to model uncertainty by importing Leapfrog surfaces and using DIPS to pull televiewer/mapping sets, then assigning shear-strength PDFs from our lab data — hands-on with LHS and sensitivity runs; about US$500 for a 1‑day session. It hit what you’re after, though 3D uncertainty depth was light (the Leapfrog→Slide2 handoff is still clunky — ), so I’d pair it with Seequent’s “Uncertainty in Geological Modelling”; details here: https://www.rocscience.com/learning/workshops.
Building on @johnny_kim87, the Seequent–Rocscience workshop in August had us import Leapfrog stochastic surfaces to Slide2 and run probabilistic LEM with televiewer DIPS sets — practical tip: correct for orientation bias before clustering and sample c–φ from lab scatter by domain instead of one global pair… It’s solid but software‑heavy; for uncertainty framing, I found Edumine’s pit slope stability module useful: https://www.edumine.com.
In the ACG masterclass in October, we drove Slide2 probabilistic runs from multiple Leapfrog realisations; the one tweak that made results believable was enforcing φ–c correlation from our triaxial scatter rather than sampling independently — “don’t let phi and c wander off on their own.” If your lab set is thin, I’ve had decent luck anchoring with GSI-derived envelopes and back‑analyzing a couple of mapped benches to tune variance…
Quick tip that paid off: in Slide2 prob runs, turn on spatial variability for c–φ with autocorrelation about 1–2 bench heights and weight televiewer/DIPS sets by kappa when sampling orientations — “sample spatially, not just globally.” Do a quick back-analysis on a real wall to scale lab-to-field shear and persistence, otherwise the failure map looks like confetti; if you need CPD credit on top, Edumine’s module is fine but lighter -.
I got the most value from Rocscience’s integration workshop this fall; the tweak that made our probabilistic Slide2 runs credible was sampling a kriged phreatic surface from piezo data alongside φ–c (matched correlation length to roughly a bench or two), which turned patchy failure probability into a coherent band tied to wet zones. If you’re light on water data, at least bracket with two pressure surfaces; details and CPD info here: https://rocscience.com/training.