The 88-kilometre power line connects Kurmuk to the Ethiopian electrical grid and is now supplying power needed to complete Kurmuk Mine commissioning and advance the site’s ramp-up. The connection includes associated substations and was completed by Ethiopian Electric Power under a 20-year power purchase agreement that provides electricity at a fixed rate of approximately US$0.04 per kilowatt-hour.

The infrastructure milestone comes as Allied Gold moves the project from construction into a series of operational readiness stages. The company said mining continues to advance, while ore stockpiles are being built to support plant feed as additional processing systems are commissioned.

Kurmuk Processing Systems Move Through Commissioning

First ore was fed through the crushing circuit earlier in September, with the handover of the circuit from the commissioning team to the operations team now under way. Allied Gold said the next stage will involve introducing ore to the grinding and downstream processing circuits as those systems are progressively completed and tested.

The company has established an ore stockpile of approximately one million tonnes and plans to increase that inventory to nearly 1.5 million tonnes, equivalent to around three months of planned ore feed ahead of full circuit commissioning. Initial mining is expected to come from the Dish Mountain and Ashashire open pits, which contain the project’s current mineral resources and reserves.

Across the site, individual systems are being completed, tested against design parameters and transferred from the construction and Kurmuk Mine commissioning team to operations. Electrical, instrumentation and other systems testing remains under way as the project progresses through its remaining commissioning activities.

Mine Moves Toward First Gold Production

Kurmuk is designed as a conventional truck-and-shovel open-pit operation with a carbon-in-leach processing plant. Allied Gold said the project contains 2.7 million ounces of Proven and Probable Mineral Reserves in the initial mining areas, with exploration continuing across the wider property to identify additional resources and extend the mine life.

The company expects to introduce first ore to the grinding circuit in the coming weeks, followed by first gold after further Kurmuk Mine commissioning is completed. Those milestones remain forward-looking and depend on the successful completion of the remaining work and testing at the site.

Allied Gold has identified Kurmuk as a major contributor to its production growth strategy, with initial full-year production guidance of 240,000 to 270,000 ounces and approximately 300,000 ounces in the following year. The mine is therefore progressing through Kurmuk Mine commissioning from infrastructure completion and processing tests toward the next stages of commissioning and eventual commercial operations.

This is strengthening the role of the aluminium recycling supply chain. More than 30 million tonnes of aluminium scrap are recycled globally each year, while the International Aluminium Institute estimates that almost 75% of all aluminium ever produced remains in use. That large installed base creates a long-term source of secondary metal as products reach the end of their useful lives.

Aluminium Recycling is Becoming a Larger Source of Metal Supply

The scale of this change can be seen in projections for post-consumer scrap. The International Aluminium Institute estimates that aluminium produced from post-consumer scrap could rise from around 22 million tonnes in 2021 to between 66 million and 80 million tonnes annually by 2050 across different scenarios. At the same time, total aluminium supply is expected to increase substantially, meaning recycled material will need to contribute alongside primary production rather than simply replace it.

This creates a stronger strategic role for secondary aluminium. Unlike primary production, which depends on extracting and refining bauxite, recycled aluminium can return material already in circulation to productive use. The energy requirement for recycling can also be up to 95% lower than producing primary aluminium, strengthening the economic and environmental rationale for expanding secondary supply.

However, the growth of recycled metal depends on more than the volume of aluminium reaching end of life. Material has to be collected, sorted, processed and returned to manufacturers in forms that meet the requirements of downstream applications. aluminium recycling supply chain development is therefore closely connected to the infrastructure supporting scrap recovery.

Post-Consumer Scrap is Expanding the Secondary Resource Base

Post-consumer scrap is particularly important because it represents aluminium recovered from products after their useful life rather than material generated during manufacturing. The International Aluminium Institute recorded around 20 million tonnes of post-consumer scrap intake in 2019, accounting for almost 60% of total scrap intake at the time. More recent scenario modelling shows the potential scale of this resource, with aluminium production from post-consumer scrap rising from 22 million tonnes in 2021 to between 66 million and 80 million tonnes by 2050, depending on the scenario.

The growing availability of this material could provide a significant source of future aluminium, but recovery rates and processing capabilities will determine how much of it becomes usable secondary metal. aluminium recycling supply chain development is consequently moving toward a more strategic role within the wider aluminium market.

The funding is being delivered through the Office of Critical Minerals and Energy Innovation (CMEI) as part of the DOE’s Mine of the Future initiative. The program seeks to create domestic testing environments where new mining technology can be demonstrated and validated under real-world conditions before moving toward broader commercial use.

The four selected projects span a range of focus areas, including mine automation, connectivity, electrification, drilling, sensing, comminution, and energy management. Each project is intended to provide physical testing infrastructure where emerging solutions can be trialed in realistic mining settings.

Four Projects Selected Under the Initiative

Innovative Wireless Technologies will build a national testbed spanning sites in Colorado, West Virginia, and Virginia. The testbed will combine underground and surface mining environments to evaluate next-generation digital, connectivity, and mine automation technologies.

The University of Arizona will develop an underground testbed in Sahuarita, Arizona, with a focus on electrification and energy-management technologies. The facility aims to support the transition of mining technology from development stages toward commercial deployment, while also targeting improvements in productivity and energy efficiency.

The Curators of the University of Missouri will create a multifunctional proving ground in Rolla, Missouri, integrating underground technology testing with advanced comminution research. This facility will also serve as a platform for workforce development related to next-generation mining innovation.

Southern Methodist University will establish what the DOE describes as the world’s first synthetic mine, located at an existing drilling and sensing test facility in Navasota, Texas. The project will offer a controlled environment for testing mining equipment, automation systems, and operational workflows before they are deployed in the field.

Broader Context for Critical Minerals Investment

The DOE says these projects are designed to accelerate the commercialization of mining technology while reducing US reliance on foreign sources of critical minerals.

“Investing in moving cutting-edge mining technologies out of the lab and into the field will help unleash America’s vast mineral resources,” US Secretary of Energy Chris Wright says. “President Trump is committed to rebuilding America’s mining industry, reducing our dependence on foreign adversaries, and securing the critical minerals that are foundational to our energy, manufacturing, and national security.”

Assistant Secretary of Energy Audrey Robertson says establishing real-world testing grounds will help accelerate technological breakthroughs across the domestic mining industry. “These investments will prove that American innovation can redefine what’s possible in modern mining,” Robertson says.

This mining technology funding follows a series of recent DOE investments in US critical mineral supply chains. Last month, the department selected seven projects for US$500 million in DOE funding to expand critical mineral processing, battery manufacturing, and recycling capacity. It also announced US$162 million for nine projects targeting the recovery of materials including scandium, copper, antimony, and rare earth elements from industrial feedstocks. Separately, DOE’s Critical Materials Innovation Hub awarded US$10 million to seven early-stage research projects focused on technologies for rare earths, gallium, copper, and other critical minerals.

This is giving AI mine planning a broader role across the mining value chain. Rather than relying solely on predefined assumptions and conventional optimisation techniques, planners can increasingly use machine learning models to process complex datasets and provide additional information about possible outcomes.

AI is Bringing More Data into Mine Planning

Geological information is one of the most important inputs. Resource estimation requires mining teams to interpret drilling, geological, geochemical and geophysical information before determining how an orebody may be extracted. Machine learning can process relationships within these datasets and assist with resource estimation or classification.

The same principle applies to operational information. Equipment utilisation, production rates, maintenance records and changing operating conditions can provide additional inputs for planning models. Bringing these datasets together can allow planners to assess proposed schedules against a wider range of real-world conditions rather than relying entirely on historical averages.

Recent research demonstrates the growing scale of interest. A 2025 bibliometric review analysed more than 1,200 publications examining artificial intelligence applications across minerals engineering, including mineral exploration, mining and mineral processing. The research identified applications spanning resource estimation, operational optimisation, safety and autonomous systems.

The growth in research does not mean that these applications have reached uniform commercial adoption. It does, however, show that AI is becoming an increasingly established area of research across the mining value chain. AI mine planning is part of this wider movement toward using more data-intensive methods to support technical and operational decisions.

From Geological Models to Data-Driven Decisions

The potential value becomes greater when AI is combined with established planning and optimisation techniques. Machine learning can provide predictions or identify relationships within complex datasets, while optimisation models can use those outputs alongside production and operational constraints.

This creates a more connected planning process. Geological uncertainty, equipment availability and production requirements can be considered together instead of being treated as completely separate analytical problems.

AI mine planning can therefore help planners work with more detailed and continuously updated information, while maintaining human oversight over final decisions. The technology is not necessarily replacing established mine-planning expertise. Its role is increasingly to expand the quantity and range of information available when evaluating alternative plans.

This is making private 5G mining increasingly relevant as mines look for connectivity infrastructure that can support multiple operational applications at the same time. Unlike conventional wireless systems designed primarily for general connectivity, private cellular networks can be configured around the coverage, performance, security and traffic requirements of a specific mine.

Connectivity is Becoming Critical Infrastructure

The need for stronger connectivity is particularly evident in large and complex mining environments. Open-pit operations can cover extensive areas, while underground mines introduce tunnels, changing layouts and difficult radio conditions. As automation expands, connectivity interruptions can affect not only communications but also remote equipment operation, machine coordination and access to real-time operational information.

Private 5G provides a dedicated network environment that can combine wide-area coverage with controlled performance. In a documented remote-dozing deployment, an existing Wi-Fi network struggled to reliably connect more than two machines at distances of around 100 metres. A private 5G deployment subsequently provided coverage across up to 2.5 kilometres from a single radio, with uplink throughput of up to 175 Mbps in the reported configuration.

The difference illustrates why private 5G mining is increasingly being considered as operational infrastructure rather than simply another wireless connectivity option. Wider coverage and greater uplink capability can support applications that depend on continuous data exchange, particularly where moving or expanding physical network infrastructure is difficult.

Connected Equipment is Raising Network Requirements

The growth of autonomous and remotely operated equipment is increasing the amount and importance of data moving across mining sites. Machines can require connectivity for control signals, telemetry, video feeds, positioning, fleet coordination and equipment monitoring. Sensors add further streams of environmental and operational information.

A single network may therefore need to support several applications with different performance requirements. Remote operation can depend on predictable latency and stable uplink capacity, while monitoring systems may generate large volumes of sensor and video data.

This broader requirement is helping position private 5G mining within the development of connected mine infrastructure. The network becomes a common layer connecting machines, sensors and computing resources rather than serving one individual application.

The industry is still transitioning, however. Global private mobile network data recorded 134 mining deployments among 1,846 private network customers across 80 countries in August 2025, showing that adoption is established but still developing. The installed base also included substantial LTE and hybrid deployments, indicating that 5G is evolving alongside existing technologies rather than replacing them immediately.

This is making mining digital twins increasingly relevant to how mines monitor assets and manage changing conditions. Instead of functioning only as static representations, advanced digital twins can combine sensor data, equipment information and operational inputs to reflect the changing state of a physical mine. This creates a digital environment that can be used to observe current conditions and support decisions based on the latest available information.

Digital Twins are Bringing Mine Operations Into Real Time

Real-time monitoring is one of the core capabilities identified in recent research on mining digital twins. Sensors and connected equipment can continuously provide information about operating conditions, equipment performance and other physical variables. That information can then be incorporated into the digital model, creating a closer relationship between what is happening in the mine and what is represented digitally.

This can improve operational visibility across areas where conditions change frequently. Equipment status, production activity and environmental conditions can all be monitored within a continuously updated digital environment. Rather than relying entirely on periodic reports or isolated measurements, operational teams can have a more current view of the state of an asset or process.

The technology also creates a basis for moving beyond observation. When real-time information is combined with analytical models, a digital twin can help identify unusual conditions and potential anomalies. This can support earlier investigation of equipment or operational issues before they develop into more significant disruptions.

Mining digital twins are therefore becoming more than visual models of mine assets. Their value increasingly comes from the connection between physical operations, continuous data collection and analytical interpretation. That connection can give mining teams a more detailed understanding of how assets and processes are behaving at a particular point in time.

From Digital Representation to Operational Intelligence

The development is also broadening the potential applications of digital twins across mining. Research identifies uses ranging from equipment monitoring and predictive maintenance to production optimisation, process simulation and environmental management. These applications share a common requirement: the digital model must remain connected to the physical operation through reliable and timely data.

This creates an important distinction between having a digital model and having a functioning digital twin. A static model can represent an asset or mine layout, but a digital twin is expected to evolve as the corresponding physical system changes. Mining digital twins consequently depend on data integration as much as modelling capability.

Key Takeaway: Mining digital twins are moving beyond static digital representation by connecting real-time monitoring with prediction, simulation and operational decision support.

As these capabilities develop, mining digital twins are becoming a mechanism for connecting operational data with a continuously updated view of the mine. The broader shift is from simply representing mining assets digitally toward creating an information environment that can support faster monitoring, analysis and operational decisions.

The $750 million investment is being made through US SIIE, LLC under the Industrial Base Analysis and Sustainment programme. It forms part of a broader $1.55 billion initiative designed to establish a secure supply chain for rare-earth elements considered critical to both national defence and economic security.

US Commits $750 Million to Serra Verde Rare Earth Project Financing

The broader $1.55 billion investment structure includes three key components. The Department of War’s Economic Defense Unit has committed $750 million. The Defense Logistics Agency has made a $300 million purchase commitment for rare-earth materials. A money-centre bank has also committed $500 million to the initiative.

Together, these commitments are intended to support an offtake agreement for mixed rare-earth carbonates produced at Serra Verde’s Pela Ema Project in central Brazil. The rare-earth elements targeted by the initiative include dysprosium, terbium, neodymium and praseodymium. These critical minerals are essential inputs for NdFeB permanent magnets, which are used across defence and industrial applications including nuclear submarines, fighter jets, satellites, guided missiles, combat vessels and drones.

The Serra Verde Rare Earth Project has previously received support through a $565 million financing agreement with the US International Development Finance Corporation. The latest $750 million commitment builds on that earlier backing and extends the financial architecture around the Pela Ema Project.

Rare-Earth Supply Chain Gains US Backing

The initiative is focused on developing a mine-to-magnet supply chain that connects rare-earth mining at the Serra Verde Rare Earth Project with emerging US magnet manufacturing capabilities. Mixed rare-earth carbonates from the Pela Ema Project would serve as feedstock for rare-earth processing and separation, ultimately supplying manufacturers producing permanent magnets for defence and industrial use.

Beyond defence, the rare-earth elements produced at Serra Verde have applications in energy infrastructure, transportation, aerospace and electronics. The breadth of end-use markets underscores the strategic value placed on building a reliable rare-earth supply chain backed by long-term offtake and purchase commitments.

The $1.55 billion investment structure around the Serra Verde Rare Earth Project reflects a concerted effort by the US government to diversify critical minerals sourcing. By anchoring supply commitments to a Brazilian rare-earth operation, the initiative aims to reduce reliance on concentrated supply sources and strengthen access to rare-earth elements for both military and civilian manufacturing.

The Serra Verde Rare Earth Project and its Pela Ema operation in central Brazil now sit at the centre of one of the most significant US-backed critical minerals financing efforts currently in progress.

That makes recovery increasingly important to mine economics. A one-percentage-point improvement can appear small on paper, but across a large operation processing millions of tonnes, the additional recovered metal can become commercially significant. The economics, however, depend on what it costs to achieve that improvement through additional energy, reagents, water, grinding, equipment or processing capacity.

Recent research shows why recovery assumptions deserve closer attention during mine planning. A 2026 copper case study from eastern Türkiye compared a fixed recovery assumption of 80% with a model that incorporated recovery variation by ore type. Under the economic cut-off scenario, the variable-recovery approach produced an average recovery of 85.8% and around 2.8 thousand tonnes of additional recovered copper. The study also found that using a fixed recovery assumption understated project NPV by roughly 8%.

The finding is important because recovery is rarely uniform across a deposit. Mineralogy, liberation, hardness, oxidation and ore type can all influence how much metal a processing plant can recover. A mine-wide average can therefore hide differences between material that is highly responsive to processing and material that is much harder to treat.

Recovery Can Add Value Without Adding More Ore

The economic logic behind recovery is straightforward: if the same amount of ore contains the same amount of metal, recovering more of that contained metal increases the amount that can potentially be sold.

The challenge is finding where the additional value exceeds the additional processing cost.

That makes recovery different from simply pushing a plant for higher throughput. More tonnes through a plant do not automatically create more value if those tonnes have lower grades or poorer metallurgical characteristics. In some cases, improving recovery from material already being processed can provide a more attractive route to additional production.

The effect can become significant at the project level. A 2025 gold-project sensitivity analysis modelled recovery between 68.3% and 85.4%. At a gold price of US$2,750 per ounce, after-tax NPV5 increased from US$1.915 billion at 68.3% recovery to US$2.761 billion at 85.4%. That is a project-specific economic sensitivity, not an industry benchmark, but it illustrates how strongly recovery assumptions can influence asset value.

For miners, the question is therefore moving beyond how much ore can be processed toward how much valuable metal can be recovered from the ore already entering the plant.

Copper Recovery from Legacy Mine Water Enters Pilot Stage

Starting in August 2026, SiTration and BHP are running an initial small-scale pilot plant at the historic Copper Cities site in Arizona’s Globe-Miami mining district. The first deployment will validate continuous and autonomous copper production from legacy mining water over a one-month period.

The Copper Cities site produced almost 400 000 tonnes of copper between the 1950s and 1980s. SiTration and BHP now see an opportunity to recover value from legacy mining assets like this one, with the potential to create new pathways for domestic US copper supply.

Prior to the pilot, SiTration conducted bench-scale testing using real feedstock from the Copper Cities site. Those tests demonstrated London Metal Exchange Grade A copper production without using any chemicals or generating new waste products. Preliminary energy consumption was reported below 4 kWh/kg to recover copper from the diluted legacy mine water.

BHP and SiTration Test Recovery Technology

The pilot deployments are designed to evaluate SiTration’s mining technology under real operating conditions. BHP legacy assets general manager Kevin Ramsay said the Copper Cities pilot provides an opportunity to assess an innovative approach to copper recovery from mining-impacted water while generating valuable technical and operational insights.

“We are excited to work with SiTration to test this technology under real operating conditions and better understand its potential to recover value from legacy mining water sources,” Ramsay said.

SiTration CEO and co-founder Brendan Smith said the American Southwest region houses billions of dollars’ worth of copper in legacy mining water. “With global copper demand projected to grow by 70% by 2050, tapping into these resources is an excellent pathway to bolster domestic supply chains while producing copper at the bottom of the global cost curve,” Smith added.

Larger Pilot Targets Copper Cathode Production

Following the initial one-month trial, a larger deployment is planned for later in 2026. That larger pilot is targeting production of up to two tonnes of commercial-scale copper cathodes over a two-month period.

The planned progression from small-scale validation to a larger trial reflects the companies’ interest in understanding how the copper recovery from legacy mine water approach performs at increasing scale. Should the pilots deliver positive results, the technology could offer a potential new route to recovering copper from legacy mining assets across the region.

Both deployments remain at the pilot and testing stage. The results will inform whether the technology can move toward broader application for copper recovery from legacy mine water sources.