Institutions, Task Automation, and the Geography of Reinstatement: Evidence and Design from Canadian Mining

This deposit provides a reproducible methods-and-evidence package that offers an institutional counterpoint to task-wise automation analysis. Task-based models of AI/automation typically proceed by measuring task substitution and inferring displacement and reinstatement from labor market outcomes. In regulated, geographically embedded sectors such as mining, task-level substitution magnitudes are often not directly observable in the public record, while governance commitments and monitoring requirements are explicit and auditable. Using a single evidence object—the KSM 2022 NI 43-101 PFS/PEA technical report—we implement a deterministic within-document portfolio test. Page-anchored “claim-units” are extracted under a frozen scope grammar that distinguishes (A) explicit automation/control commitments from (B) explicit employment/legitimacy commitments; scope assignment is audited via search logging and disciplined by document-internal falsifiers. A further wedge is installed sensing and control capacity (IoT). AI effectiveness is frequently conditional on an accumulated instrumentation/telemetry/control layer, making AI adoption an institutionally moderated investment process into measurement-and-control capacity rather than a pure task-substitution shock. The deposit therefore includes a deterministic IoT Capacity Index protocol based on mine-plan instrumentation references. The package also includes a design instrument, Historical Episode Family Simulation (HEFS), which constructs a deterministic family of long-horizon copper price regimes and maps monthly prices into annual mine revenue under a locked aggregation contract. HEFS is intended for transparent regime stress testing (not forecasting). Concept DOI (all versions): 10.5281/zenodo.18440410Version DOI (this release): 10.5281/zenodo.18824574