What is the Aclara Penco Module AI rare earth exploration and how does it work?

Introduction to the Aclara Penco Module AI Rare Earth Exploration System

The Aclara Penco Module AI rare earth exploration system represents a specialized application of artificial intelligence in the domain of heavy rare earth element (HREE) discovery, specifically deployed in the Penco region of Chile. Aclara Resources, a Canadian-based critical minerals company, has developed this AI-driven exploration platform as part of its broader strategy to identify and develop non-brine, clay-hosted heavy rare earth deposits. The Penco Module is not a standalone software product but rather an integrated geological intelligence framework that combines machine learning algorithms, hyperspectral imaging, historical drilling databases, and geochemical assay datasets to generate probabilistic targets for follow-up fieldwork. As of September 2026, the module has been validated through multiple drilling campaigns in the Penco basin, with results indicating a significant improvement in discovery efficiency compared to traditional exploration methods.

Also worth reading: How does AI copper exploration targeting work and what makes it effective for finding new deposits in 2026? · What are autonomous critical mineral exploration platforms and how do they work in 2026? · How are quantum gravity sensors changing mineral exploration for rare earths in 2026?

The system operates on the premise that heavy rare earth elements in the Penco region are hosted in ion-adsorption clays formed through in-situ weathering of granitic parent rocks. These deposits are geologically distinct from the more commonly known Chinese ion-adsorption clays, which are typically enriched in medium rare earths. The Penco deposits are characterized by higher concentrations of dysprosium and terbium—two elements critical for high-performance permanent magnets used in electric vehicles and wind turbines. The AI module ingests over 400,000 data points from legacy soil surveys, airborne geophysical surveys (including magnetic, radiometric, and electromagnetic datasets), and satellite-derived spectral indices. These inputs are processed through a ensemble of random forest classifiers, gradient boosting machines, and convolutional neural networks trained on known mineralized zones within the Penco basin.

The architecture of the Aclara Penco Module is designed for iterative refinement. Each new drilling campaign feeds back into the model, reducing uncertainty in the predictive maps and improving the spatial resolution of target zones. The system outputs probability maps with confidence intervals, georeferenced to within 2-meter accuracy, which are then used by Aclara's exploration team to prioritize drill hole locations. The economic implications are substantial: by reducing the number of dry holes (drill holes that fail to intersect mineralization), the module lowers exploration capital expenditure by an estimated 35% compared to conventional grid-based drilling approaches. This efficiency gain is particularly important in Chile, where environmental permitting timelines can extend exploration cycles by 18–24 months.

Technical Architecture and Data Inputs

The Aclara Penco Module AI system is built on a multi-layered data ingestion pipeline that integrates terrestrial, aerial, and orbital data sources. At the base layer, high-resolution topographic data from LiDAR surveys conducted in 2023 and 2024 provides microgeomorphological insights, identifying slope breaks and drainage patterns that correlate with clay accumulation zones. This is overlaid with radiometric data from airborne gamma-ray spectrometry, which measures potassium, uranium, and thorium anomalies—proxy indicators for rare earth element enrichment. The spectrometry data is processed using a proprietary normalization algorithm that accounts for soil moisture variations and vegetative cover, both of which can mask radiometric signatures in the coastal Penco environment.

The second layer consists of hyperspectral imagery captured by the PRISMA satellite (operated by the Italian Space Agency) and the upcoming EnMAP mission. These sensors operate in the short-wave infrared (SWIR) range, where specific absorption features of rare earth element-bearing minerals such as bastnäsite and monazite are detectable. The AI module applies a spectral unmixing algorithm to deconvolve mixed pixel signatures, isolating the spectral fingerprint of rare earth mineral assemblages. This is particularly challenging in the Penco region due to the presence of iron oxides and clay minerals that produce overlapping spectral features. The neural network component of the module was trained on 1,200 hand-held spectral measurements collected during the 2025 field campaign, achieving a classification accuracy of 87% for rare earth mineral identification.

Geochemical data forms the third pillar of the system. Aclara has compiled a database of 2,300 soil samples analyzed via ICP-MS (inductively coupled plasma mass spectrometry) for 45 elements. These samples are spatially distributed across the Penco basin, with a focus on the B-horizon (subsurface soil layer) where rare earth elements typically concentrate. The AI module uses these assays as training labels for supervised learning models. The feature engineering process includes creating derived variables such as the cerium anomaly (Ce/Ce), europium anomaly (Eu/Eu), and the sum of heavy rare earths (HREE sum), which are known geochemical indicators of mineralization potential. The final output is a series of probabilistic maps showing the likelihood of encountering economic-grade HREE concentrations at depths between 5 and 30 meters.

Validation and Performance Metrics

The Aclara Penco Module has undergone rigorous validation through a series of blind prediction exercises. In the first phase (January–March 2026), the module was used to predict the locations of 15 drill holes in an area of the Penco basin that had not been previously explored. The predictions were based solely on AI-generated probability maps, with no input from human geologists. Of the 15 drill holes, 11 intersected mineralized zones with average HREE grades of 1,200 ppm total rare earth oxide (TREO), of which 28% was heavy rare earths (dysprosium plus terbium). This represents a 73% success rate, compared to the industry average of 35–40% for greenfield rare earth exploration.

In the second validation phase (June–August 2026), the module was tested against traditional exploration methods in a side-by-side comparison. A grid-based drilling approach was deployed in a neighboring concession, covering the same geological formation but with a different exploration strategy. The grid approach required 42 drill holes to identify 5 mineralized zones, yielding a success rate of 12%. The AI module, in contrast, identified 8 mineralized zones using only 18 drill holes, achieving a success rate of 44%. The economic implications are significant: the AI-driven approach reduced drilling costs by approximately USD 1.8 million (assuming USD 150 per meter for drilling in the Penco region) while discovering 60% more mineralized zones.

The performance metrics extend beyond simple success rates. The AI module demonstrates superior precision in identifying the spatial extent of mineralization. In follow-up drilling around the initial discoveries, the module's probability maps guided the placement of infill holes with an average spacing of 50 meters, compared to the 100–200 meter spacing typical of conventional exploration. This higher resolution enables more accurate resource estimation, reducing the classification uncertainty in the inferred resource category. According to Aclara's internal estimates, the use of the Penco Module could reduce the time required to reach a preliminary economic assessment (PEA) from 36 months to 24 months, representing a 33% acceleration in the project development timeline.

Practical Implementation Steps for Exploration Teams

Implementing the Aclara Penco Module requires a structured approach that balances technological capabilities with ground-truth validation. The first step involves data acquisition and preprocessing. Exploration teams must secure access to the full suite of input datasets, including LiDAR elevation models, airborne geophysical surveys, and satellite imagery. For the Penco region, Aclara has negotiated data-sharing agreements with the Chilean geological survey (SERNAGEOMIN) and the European Space Agency, ensuring that the latest available imagery is incorporated into the model. Data preprocessing includes atmospheric correction for satellite imagery, Kriging interpolation for geophysical data gaps, and quality control filtering for geochemical assays.

The second step involves model calibration and local adaptation. While the core algorithms of the Penco Module are pre-trained on the Penco basin dataset, teams operating in adjacent regions must recalibrate the model using local training data. This process, known as transfer learning, involves fine-tuning the neural network weights on a smaller dataset of local assays and mineralized intercepts. Aclara provides a standardized calibration protocol that requires a minimum of 50 validated drill holes to achieve reliable predictions. The protocol includes a spatial cross-validation procedure to ensure that the model does not overfit to local anomalies that may not be representative of the broader geological setting.

The third step is target prioritization and field deployment. The AI module generates a ranked list of exploration targets based on a composite score that integrates probability of mineralization, predicted grade, overburden thickness, and accessibility. Targets with scores above 0.75 (on a 0–1 scale) are flagged for immediate follow-up. Field teams use these rankings to allocate drilling budgets efficiently. In practice, this means that high-priority targets receive closer-spaced drilling (25–50 meter intervals), while lower-priority targets may be tested with wider-spaced drilling (100–200 meters) or deferred to later exploration phases. The system also includes a feedback loop: results from new drilling are automatically incorporated into the model, triggering a re-ranking of remaining targets within 48 hours of assay receipt.

Comparison with Alternative Exploration Approaches

The Aclara Penco Module AI system can be compared to several alternative approaches for rare earth exploration, each with distinct advantages and limitations. The table below summarizes the key differences:

Traditional grid drilling, while conceptually simple, suffers from low discovery rates and high environmental impact. The systematic grid approach often results in large numbers of dry holes, particularly in areas with complex geological structures. Manual geological mapping, though valuable for initial reconnaissance, lacks the spatial resolution necessary for identifying clay-hosted rare earth deposits, which can be highly localized. UAV-based geophysical surveys offer improved resolution compared to airborne methods but are limited by battery life and regulatory restrictions in remote areas.

The Aclara Penco Module's primary advantage lies in its ability to integrate diverse data sources and identify subtle patterns that would be imperceptible to human analysts. For example, the module can detect correlations between radiometric anomalies and specific soil geochemical signatures that are not apparent through simple threshold-based analysis. This is particularly important in the Penco region, where rare earth mineralization is often associated with subtle variations in clay mineralogy and iron oxide content that can be masked by background geochemical noise.

Common Pitfalls and Mitigation Strategies

Several common pitfalls can undermine the effectiveness of the Aclara Penco Module. The first is over-reliance on AI predictions without sufficient ground-truth validation. While the module achieves high accuracy in mineralized zone identification, it is not infallible. False positives can arise from geological features that mimic rare earth signatures, such as phosphorite nodules or ironstone concretions. To mitigate this, Aclara recommends a staged validation approach: initial drill holes should be spaced at 100-meter intervals to confirm the lateral continuity of mineralization before committing to closer-spaced drilling.

The second pitfall is inadequate data quality control. The AI module's performance is directly dependent on the accuracy and completeness of input data. Common issues include: (1) satellite imagery affected by cloud cover or atmospheric haze, (2) radiometric surveys with inconsistent calibration across flight lines, and (3) geochemical assays with detection limits that vary between laboratories. Aclara addresses these issues through a rigorous data validation protocol that includes cross-checking satellite-derived spectral indices with field-based measurements, recalibrating radiometric data using ground control points, and requiring duplicate assays from independent laboratories for 10% of all samples.

The third pitfall is neglecting the geological context. The AI module, while powerful, operates as a pattern recognition tool that may not fully account for the three-dimensional complexity of mineralization. In the Penco region, rare earth deposits can be highly discontinuous due to faulting, variations in parent rock composition, and post-mineralization weathering. To address this, Aclara incorporates geological constraints into the model by masking out areas with known faults or lithological contacts that are unfavorable for clay-hosted mineralization. This geological filtering reduces the number of false positives and improves the overall efficiency of the exploration program.

Cost-Benefit Analysis and Economic Considerations

The economic implications of the Aclara Penco Module extend beyond simple cost savings in exploration. The module's ability to accelerate discovery has significant downstream effects on project valuation and investment attractiveness. According to a preliminary economic assessment conducted in July 2026, the use of the AI module could reduce the pre-production capital expenditure for a Penco-style deposit by approximately 22%, from USD 450 million to USD 351 million. This reduction is primarily driven by more accurate resource estimation, which minimizes the need for excessive drilling to define ore reserves.

The cost structure of the AI module itself is worth examining. Aclara charges a licensing fee of USD 250,000 per year for access to the proprietary algorithms and data processing infrastructure. This includes ongoing model updates, technical support, and access to the cloud-based computing platform. For exploration companies with limited budgets, Aclara offers a pay-per-use model at USD 5,000 per target zone analyzed, with a minimum commitment of 10 zones. The total cost of implementing the module, including data acquisition and field validation, typically ranges from USD 750,000 to USD 1.2 million for a comprehensive exploration program covering 50–100 square kilometers.

The return on investment (ROI) for the AI module is particularly compelling when considering the opportunity cost of delayed discovery. In the current market environment, where rare earth element prices are volatile and supply chain security is a growing concern, every month of accelerated discovery can translate to millions of dollars in market capitalization gains. For example, Aclara's stock price increased by 5.88% in a single trading session following the announcement of successful AI-guided drilling results, representing a market valuation increase of approximately USD 45 million. This suggests that the AI module's economic benefits extend well beyond direct exploration cost savings.

Future Outlook and Integration with Broader Supply Chain Initiatives

Looking ahead, the Aclara Penco Module is poised to play a central role in Aclara's broader strategy to develop a vertically integrated heavy rare earth supply chain. The company has partnered with Argonne National Laboratory to develop an AI-enabled digital twin for heavy rare earth separation, which will directly incorporate the geological models generated by the Penco Module. This integration will enable seamless data flow from exploration through to processing optimization, potentially reducing the overall development timeline by an additional 12–18 months.

The U.S. Department of Energy's decision to select Aclara for federal funding to advance AI-driven heavy rare earth processing further underscores the strategic importance of the Penco Module. The DOE grant, valued at USD 15 million, will support the development of machine learning models for optimizing separation chemistries and reducing reagent consumption in the processing plant. These models will be trained on the same geological datasets used by the Penco Module, creating a unified AI framework that spans the entire value chain.

From a market perspective, the Penco Module's capabilities are becoming increasingly relevant as Western countries seek to reduce their dependence on Chinese rare earth supplies. China currently controls approximately 60% of global rare earth production and 85% of heavy rare earth processing capacity. The Penco deposits, with their high concentrations of dysprosium and terbium, represent one of the few large-scale alternative sources of these critical elements outside of China. The AI module's ability to accelerate discovery and reduce exploration risk is therefore not just a technical achievement but a strategic imperative for supply chain diversification.

Frequently Asked Questions

What makes the Aclara Penco Module different from other AI exploration tools? The Penco Module is specifically optimized for clay-hosted heavy rare earth deposits in the Penco basin, incorporating region-specific geological knowledge and training data. Unlike generic AI exploration platforms, it integrates hyperspectral imaging, radiometric data, and geochemical assays in a unified framework that accounts for the unique weathering and mineralization processes in this part of Chile.

How accurate are the predictions generated by the AI module? The module achieves a 73% success rate in blind prediction exercises, compared to industry averages of 35–40% for rare earth exploration. The accuracy is measured by the percentage of drill holes that intersect mineralized zones with economic-grade rare earth concentrations. The module also provides confidence intervals for each prediction, allowing exploration teams to assess risk levels.

What is the minimum data requirement for implementing the Penco Module? The module requires access to LiDAR elevation data, airborne geophysical surveys (magnetic, radiometric, electromagnetic), satellite hyperspectral imagery, and a minimum of 50 validated geochemical assays for calibration. These datasets are available through Aclara's data partnerships or can be acquired independently at an estimated cost of USD 200,000–400,000.

Can the Penco Module be applied to other rare earth deposit types? While the core algorithms are transferable, the module is specifically calibrated for ion-adsorption clay deposits. Application to other deposit types (e.g., carbonatite-hosted or placer deposits) would require significant recalibration and additional training data. Aclara offers customized versions of the module for different geological settings upon request.

What are the environmental implications of AI-guided exploration? The AI module reduces environmental impact by minimizing the number of drill holes required for discovery. In the Penco validation study, the AI approach required 57% fewer drill holes to achieve the same number of discoveries as conventional methods. This translates to reduced land disturbance, lower fuel consumption, and shorter permitting timelines.

Quick Facts

## Sources

• Aclara Resources press release on technical report results (Newswire.com, September 2026)
• Aclara and Argonne National Laboratory partnership announcement (Investing News Network, August 2026)
• U.S. Department of Energy funding selection (ACCESS Newswire, July 2026)
• CEO Letter to Shareholders: 2025 Accomplishments and 2026 Outlook (Newswire.com, January 2026)
• Kalkine.ca analysis of Aclara stock performance (September 2026)
• Proactive Financial News on Chilean environmental review progress (August 2026)
• Discovery Alert on Aclara heavy rare earth facility (July 2026)
• Investing News Network: 4 Best-performing Canadian Rare Earths Stocks in 2026 (September 2026)

Follow-up Keyword

Aclara Penco AI rare earth discovery efficiency