Research
The Intelligent Mine Is Being Built Now. The Operating Model Is Not.
The mining sector is deploying agentic AI, generative AI, and advanced robotics across exploration, operations, processing, and logistics. The technology is transforming what mining can do. But the way operations are organised, controlled, and governed has not kept pace, and the gap between what the technology makes possible and what mining organisations actually achieve is widening, not closing.
A fully realised intelligent mining operation is not an incremental improvement on what exists today. It is a fundamentally different system: one in which AI continuously processes operational data across the entire value chain, autonomous systems act without moment-to-moment human instruction, distributed intelligence integrates what were previously separate functions; planning, maintenance, geology, safety, logistics into a single decision environment, and the control and decision centre is the operational brain of the whole, not a remote monitoring room.
Most mining operations are not building toward that system. They are bolting automation onto operating models designed for an earlier era of manual operation and physical presence, and wondering why the returns are below expectation. The reason is not the technology. It is the operating model.
This research addresses that problem directly.
The Six Research Questions
Creating a New Operating Model for Intelligent Mining Operations
Human Authority-Centred Work Design for Intelligent Mining Operations
Principal Investigator: Dr Peta Chirgwin, Founder & CEO, Chameleon Mettle Group
This research program is structured around one primary question and six supporting sub-questions, combining technology analysis, operational fieldwork, human factors research, organisational design, and economic modelling to produce both rigorous academic contributions and directly applicable industry frameworks.
The central proposition: the control and decision centre, properly conceived, is not a place where people watch machines work. It is the operational hub of an intelligent system, the point at which human judgement and artificial intelligence meet, where distributed operational data becomes strategic decision-making, where autonomous systems are directed and interrogated, and where the mining company's most consequential operational decisions are made. Designing that centre and the operating model around it is what this research does.
SQ1 — What does a fully realised intelligent mining operation look like?
Before designing the control and decision centre, the research establishes the technological and operational canvas: where AI, robotics, and distributed intelligence are already transforming mining, where deployment remains aspirational, and what an integrated intelligent mining operation looks like when these systems work together rather than being deployed piecemeal. This is the foundational question. Without a clear answer, the design questions that follow lack the operational context they require.
SQ2 — What is the architecture of a next-generation control and decision centre?
This is the central design question. The research examines what a control centre must do not just what it must contain when it is designed as the hub of an intelligent system rather than a remote monitoring facility. It distinguishes between passive surveillance (the dominant current model) and active decision-making (the model this research proposes), and examines how that distinction drives fundamentally different choices about physical environment, technology architecture, interface design, and decision protocols. Key questions include how situational awareness is constructed across a distributed intelligent system, how AI-generated intelligence is surfaced without inducing cognitive overload or automation bias, and how the centre remains functional when components of the distributed system fail.
SQ4 — What new roles and organisational structures does an intelligent mining operation require?
This sub-question designs the human organisation of the intelligent mining operation from first principles. It moves beyond the question of which jobs will be lost to examine what genuinely new work emerges and what work must be deliberately preserved. It examines how roles should be redesigned around human comparative advantage in an AI-augmented environment, how organisational structures need to change when intelligence is distributed and decision-making is concentrated in a central hub, and what happens to the tacit operational knowledge experienced miners carry when it is no longer exercised in a field workforce. CMG's Human Authority Index (HAI™) provides the methodology for specifying, within role and system design, where human judgement and authority must be intentionally preserved rather than delegated to automated systems.
SQ5 — What human factors principles, safety frameworks, and regulatory structures are required?
This sub-question stress-tests the operating model against the conditions that make intelligent mining viable rather than merely possible. It examines how people and intelligent systems must be jointly designed to work together effectively, drawing on sociotechnical systems theory, cognitive ergonomics, and human factors research from aviation, offshore energy, and nuclear operations — industries that have already navigated the transition to complex partially-autonomous systems. It then maps the regulatory gap: current WA and Australian mining regulation was designed for human operators and mechanical equipment and is largely silent on AI-influenced decisions, autonomous systems, and distributed accountability. When a distributed AI system contributes to a safety incident, how is responsibility attributed? This sub-question identifies where regulatory frameworks must evolve and what governance architectures from analogous industries offer as models.
SQ6 — Does this operating model give mid-tier mining companies a genuine competitive advantage?
This sub-question examines the strategic and economic case. Mid-tier WA mining companies face a distinctive version of the intelligent operations challenge: many are approaching their first serious investment in advanced operations without the legacy systems, entrenched structures, or sunk technology decisions that constrain Tier 1 operators. This gives them a rare opportunity — to design from first principles, building an intelligent operation that is integrated by design rather than integrated by retrofit. The research tests whether that structural difference produces a meaningfully different cost profile, safety profile, and performance trajectory, and under what conditions the advantage holds. It produces an economic model and strategic framework directly usable by boards, investors, and company decision-makers contemplating first entry into advanced operations.
Why This Research Matters for WA
Beyond the operational case, this research addresses a broader strategic question that WA's mining sector has not yet confronted directly.
Intelligent systems create genuine productivity gains. But where those systems are owned, continuously improved, and licensed by entities based overseas, a portion of the value those gains represent wages, tax revenue, and the tacit knowledge of experienced workers converts into a licence fee paid offshore. Australia has long sent its critical minerals overseas to be processed, capturing less of the value chain than its resource base warrants. This research argues that operational data and the intelligence derived from it are now a comparable resource, and that the choices WA's mining sector makes in the next decade about how intelligent systems are procured, governed, and integrated into its workforce will determine whether it captures or exports that value.
The research produces a Value Retention Index; a methodology for measuring how much of the value created by intelligent systems stays in the Australian economy and a redeployment-versus-redundancy decision framework that gives operators and policymakers an evidence base for how productivity gains from intelligent systems are captured.
Join the Research
This research program is currently in the collaboration-building phase. CMG is seeking industry sponsors, pilot site partners, and research collaborators to join the program ahead of a formal funding application.
Early involvement matters. Sponsors and collaborators who join at this stage have direct input into the research design — shaping the sub-questions, the fieldwork methodology, and the design of outputs so they are directly applicable to their operational context. First-mover collaborators are named as founding partners in the funding application and all subsequent outputs.
Industry Sponsors provide co-funding and receive first-access rights to all research outputs before public release — including the Control and Decision Centre Design Framework, Value Retention Index methodology, HAI™ assessment tool, role specifications, and regulatory guidance. Sponsors also participate directly in co-design workshops and have their operational context embedded in the research design.
Pilot Site Partners provide site access — control room observation, role mapping, and personnel time for HAI™ assessment — and receive in return a site-specific HAI™ baseline report and early access to HACWD-aligned role and interface design recommendations for their operations. Mid-tier operators designing their first advanced operations are particularly encouraged to engage.
Research Collaborators — universities, METS sector organisations, industry associations, and government agencies — contribute specialist expertise, fieldwork access, or sector-wide data across one or more of the six sub-questions, and are acknowledged as named partners in all research outputs and publications.
All partnership arrangements are structured around clear IP terms: CMG retains ownership of the HAI™ methodology and HACWD framework; co-developed outputs are shared under agreed licence arrangements; and company-specific data is fully de-identified in all published outputs.
Express Interest
If you are a mining operator, technology provider, industry association, or research institution interested in participating in this program as a sponsor, pilot site partner, or research collaborator, we welcome a conversation.
This website uses cookies.
We use cookies to analyze website traffic and optimize your website experience. By accepting our use of cookies, your data will be aggregated with all other user data.