Track 1: AI and Data-Driven Decision Making

DATA-DRIVEN DECISION MAKING AND VALUE CHAIN OPTIMIZATION IN UNDERGROUND MINING: A DESWIK MDM CASE STUDY AT NEXA RESOURCES ARIPUANÃ *V.R.N. Cordeiro1, W.B.R. Espejo1 1Nexa Resources, S.A., (*Presenting author: viniciusrewel@gmail.com) ABSTRACT Information fragmentation across heterogeneous systems, discipline-specific repositories, and uncontrolled file versions is a persistent operational challenge in underground mining, directly affecting planning reliability, inter-departmental alignment, and—critically—process safety. In environments where analyses performed on outdated designs can propagate hazardous assumptions into field execution, data governance is not merely a productivity concern but a riskcontrol imperative. This paper presents the implementation of Deswik Mining Data Management (MDM) at Nexa Resources' Aripuãnã underground mine (Mato Grosso, Brazil), one of South America's most complex polymetallic operations (Cu–Zn–Pb). We deployed the platform as a governed, single source of truth encompassing four interconnected layers: (i) standardized data ingestion from heterogeneous sources; (ii) automated validation and version control; (iii) structured workflow automation; and (iv) Business Intelligence (BI) integration for real-time decision support. Our initiative engaged 80+ users across six technical disciplines—mine planning (short-, medium-, and long-term), geology, geomechanics, topography, ventilation, and infrastructure—over a six-month deployment period, with more than 40 standardized workflows configured to govern data flows, enforce multidisciplinary validation gates, and automate routine reconciliation tasks. Quantitative outcomes, measured against a pre-implementation baseline, demonstrate: (a) ~90% reduction in linear advance measurement cycle time; (b) ~80% reduction in stoping reconciliation cycle time; and (c) ~60% reduction in drilled reserve reconciliation errors through spatially governed, audit-ready processes. Beyond efficiency gains, we eliminated parallel file versions, reducing the probability of field decisions based on superseded technical information and functioning as a measurable process safety assurance mechanism. This case study demonstrates how we positioned MDM-driven data governance as the foundational enabler for Short Interval Control (SIC), integrated control-room operations, and closed-loop decisionmaking—key pillars of an Integrated Mining Enterprise (IME) under the Mining 4.0 paradigm. The results establish a replicable governance architecture that accelerates digital maturity progression, quantifiably improves safety assurance through data governance discipline, and unlocks competitive advantage in increasingly complex polymetallic mining environments. KEYWORDS

RkJQdWJsaXNoZXIy MTM0Mzk2