Distributionally Robust Scheduling of Geological Engineering Systems in Uganda: Reliability Guarantees Under Demand and Failure Uncertainty

Scheduling geological engineering systems in data-scarce regions such as Uganda requires balancing operational efficiency against the risk of unmet demand and cascading failure. Conventional stochastic programming assumes a known probability distribution for uncertain parameters, an assumption that is rarely defensible when hydrological, geotechnical and demand data are sparse. The framework integrates a Markovian failure model with a Wasserstein-metric ambiguity set, yielding a tractable convex reformulation that preserves the structure of the underlying scheduling problem. Formal results establish finite-sample reliability guarantees, demonstrating that the proposed approach outperforms sample-average approximation when historical data are limited. A methodological workflow for Ugandan geological engineering contexts is presented, linking the mathematical framework to spatial data infrastructure and decision-making under uncertainty. The analysis shows that distributional robustness offers a principled alternative to both deterministic and stochastic scheduling when distributional knowledge is incomplete, and provides a template for reliability-constrained infrastructure planning in analogous data-limited environments.