The Top-Cut
Geology, mining and the machine learning creeping into resource estimation.
Listen in your podcast app
Episodes
Special — Benchmarking the Mine Plan
A special episode. Traci and Joao take the reference-class idea from Flyvbjerg and Gardner's How Big Things Get Done and apply it to the equipment assumptions inside mine plans: availability and how it decays with age, utilisation, maintenance cost versus age, replacement timing, nameplate versus achieved productivity, what automation really buys, and truck-shovel fleet sizing and scheduling. Drawing on about forty sources, from Luleå theses and CIM, AusIMM and SME papers back to 1983 through to vendor handbooks and fleet-dispatch studies, they build ten carry-forward rules of thumb with ranges, sources and conditions. They then test them against the Era Dorada, Las Chispas, Cariboo and Platreef technical reports. Every number was checked against its source, and unread papers are flagged as such.
Ep. 1 — AI Meets the Block Model
Traci and Joao work through a year of machine learning applied to mineral resource and reserve estimation: whether anything actually beats kriging, the geology-as-feature versus geology-as-constraint distinction, generative models moving into conditional simulation, uncertainty a Competent Person could defend, and a ranked reading list. Roughly twenty papers, every citation verified against a primary source.
Both hosts are synthetic voices and the script is AI-generated. Every paper cited was verified against a primary source (DOI or arXiv ID) before publication, but the commentary is not that of practising researchers.