Periodic Reporting for period 1 - PERSEPHONE (Autonomous Exploration and Extraction of Deep Mineral Deposits)
Okres sprawozdawczy: 2024-01-01 do 2025-06-30
The world is highly dependent on imported critical raw materials (CRMs), many of which are essential for clean energy, mobility, electronics, and defence. With rising global demand, supply instability, and increasing societal and environmental expectations, there is an urgent need to develop innovative and sustainable ways to access Europe’s own domestic raw material resources. However, many of these resources lie in deep, remote, or previously abandoned underground mines, where conventional mining is either unsafe, economically unviable, or ecologically damaging. This is the exact focus of the PERSEPHONE project that aims in developing novel technologies to reduce human risk, minimize surface disruption, and enable precise, data-driven operations.
In more detail, PERSEPHONE sets out to develop a new generation of autonomous exploration and extraction technologies that can operate without human presence in hazardous underground settings. The project combines advances in robotics, sensing, artificial intelligence, and digital twin technologies to enable precise drilling, continuous rock characterization, and computer-aided mine planning. By integrating these capabilities, PERSEPHONE aims to reduce energy consumption, improve ore recovery, and minimise environmental impact.
To achieve this, the project had defined five Specific Objectives (SOs) that together shape the technical, operational, and strategic goals of the initiative:
SO1: Develop independent and autonomous near-mine exploration and extraction mining machines capable of operating in unstructured, confined environments without relying on external infrastructure.
SO2: Enable efficient and accurate autonomous near-mine exploration and extraction of mineralisations, including sensing, sampling, and orebody characterization.
SO3: Develop dynamic, digital twin–based and interoperable geo-models to support optimal planning of face drilling and extraction workflows.
SO4: Enable online, in-situ analytics technologies for ore grading while drilling, which supports data-driven, adaptive mine planning and extraction systems that connect exploration insights with production planning and real-time feedback loops.
SO5: PERSEPHONE concept validation in deep and abandoned mines, that ensure the integration of the developed robot designs and validation of autonomous mining systems, contributing to responsible innovation.
The project is closely aligned with EU policy priorities, including the EU Green Deal, the Critical Raw Materials Act, and the Raw Materials Initiative. It directly supports the goal of enhancing strategic autonomy and ensuring that Europe’s green and digital transitions are not jeopardized by resource dependency.
In more detail, PERSEPHONE sets out to develop a new generation of autonomous exploration and extraction technologies that can operate without human presence in hazardous underground settings. The project combines advances in robotics, sensing, artificial intelligence, and digital twin technologies to enable precise drilling, continuous rock characterization, and computer-aided mine planning. By integrating these capabilities, PERSEPHONE aims to reduce energy consumption, improve ore recovery, and minimise environmental impact.
To achieve this, the project had defined five Specific Objectives (SOs) that together shape the technical, operational, and strategic goals of the initiative:
SO1: Develop independent and autonomous near-mine exploration and extraction mining machines capable of operating in unstructured, confined environments without relying on external infrastructure.
SO2: Enable efficient and accurate autonomous near-mine exploration and extraction of mineralisations, including sensing, sampling, and orebody characterization.
SO3: Develop dynamic, digital twin–based and interoperable geo-models to support optimal planning of face drilling and extraction workflows.
SO4: Enable online, in-situ analytics technologies for ore grading while drilling, which supports data-driven, adaptive mine planning and extraction systems that connect exploration insights with production planning and real-time feedback loops.
SO5: PERSEPHONE concept validation in deep and abandoned mines, that ensure the integration of the developed robot designs and validation of autonomous mining systems, contributing to responsible innovation.
The project is closely aligned with EU policy priorities, including the EU Green Deal, the Critical Raw Materials Act, and the Raw Materials Initiative. It directly supports the goal of enhancing strategic autonomy and ensuring that Europe’s green and digital transitions are not jeopardized by resource dependency.
In the first half of the project (Months 1–18), PERSEPHONE has made significant progress in the development, testing, and integration of its core technologies:
• Novel Autonomous robotic platforms (images below) have been developed and tested in simulations and laboratory conditions, thus being field ready for deployment. These include modular drilling robots and mobile exploration units are capable of navigating and operating without external infrastructure.
• Advanced state of the art sensor systems are being validated, including Laser-Induced Breakdown Spectroscopy (LIBS), multispectral imaging, and Measurement While Drilling (MWD) tools. These systems enable real-time identification of ore quality and lithological boundaries during drilling and exploration.
• A flexible and modular autonomy stack has been created to support fully autonomous mining machines, including software for risk-aware navigation, multi-robot coordination, and precision alignment for drilling tasks. This system integrates inputs from LiDAR, radar, cameras, and IMUs to plan and execute complex underground operations, which have been tested in the simulation environments.
• A dynamic geo-modelling pipeline has been developed and tested in simulations to support the continuous updating of digital mine models based on the sensing data. These models feed into digital twin platforms for planning, visualization, and reporting.
• The project has developed the simulation environments (using Gazebo and ROS) that replicate real mine topologies using 3D point clouds and mesh models. These virtual mines allow for iterative testing, validation, and reproducibility.
• Novel Autonomous robotic platforms (images below) have been developed and tested in simulations and laboratory conditions, thus being field ready for deployment. These include modular drilling robots and mobile exploration units are capable of navigating and operating without external infrastructure.
• Advanced state of the art sensor systems are being validated, including Laser-Induced Breakdown Spectroscopy (LIBS), multispectral imaging, and Measurement While Drilling (MWD) tools. These systems enable real-time identification of ore quality and lithological boundaries during drilling and exploration.
• A flexible and modular autonomy stack has been created to support fully autonomous mining machines, including software for risk-aware navigation, multi-robot coordination, and precision alignment for drilling tasks. This system integrates inputs from LiDAR, radar, cameras, and IMUs to plan and execute complex underground operations, which have been tested in the simulation environments.
• A dynamic geo-modelling pipeline has been developed and tested in simulations to support the continuous updating of digital mine models based on the sensing data. These models feed into digital twin platforms for planning, visualization, and reporting.
• The project has developed the simulation environments (using Gazebo and ROS) that replicate real mine topologies using 3D point clouds and mesh models. These virtual mines allow for iterative testing, validation, and reproducibility.
PERSEPHONE has already achieved the following results that go beyond the state of the art in mining and digitalization:
• Underground modular mining machine fleet: Project has achieved the first models of conceptual design of novel modular mining machines for collaborative operation in deep and abandoned mines that has been tested in simulated environments developed in the project from the real mining data.
• Zero-human presence mining: The project has demonstrated in simulations the feasibility of operating mining machines underground without any reliance on existing infrastructure such as GPS or Wi-Fi.
• Collaborative machine autonomy: Multi-machine coordination, navigation and map merging algorithms validated in simulation established the foundation for scalable machine fleets in future mining operations.
• AI-enabled rock characterization: Project achieved successful real-time data integration from MWD, spectral and multispectral sensing data with geological modelling tools.
• Digital twin integration: Dynamic updating of geological block models based on live synthetic sensor streams has been achieved in the project.
Looking ahead, the project keeps its ambition towards achieving its larger impacts and to ensure further marker uptake. These activities will be implemented between M19 and M36 and will include extended field demonstrations in operational mines, further industrial validation of the robotic fleet, and the establishment of standards for adaptive and data-driven digital twin integration and mine planning that will serve as a unified tool for all the operational categories and financial decision making.
• Underground modular mining machine fleet: Project has achieved the first models of conceptual design of novel modular mining machines for collaborative operation in deep and abandoned mines that has been tested in simulated environments developed in the project from the real mining data.
• Zero-human presence mining: The project has demonstrated in simulations the feasibility of operating mining machines underground without any reliance on existing infrastructure such as GPS or Wi-Fi.
• Collaborative machine autonomy: Multi-machine coordination, navigation and map merging algorithms validated in simulation established the foundation for scalable machine fleets in future mining operations.
• AI-enabled rock characterization: Project achieved successful real-time data integration from MWD, spectral and multispectral sensing data with geological modelling tools.
• Digital twin integration: Dynamic updating of geological block models based on live synthetic sensor streams has been achieved in the project.
Looking ahead, the project keeps its ambition towards achieving its larger impacts and to ensure further marker uptake. These activities will be implemented between M19 and M36 and will include extended field demonstrations in operational mines, further industrial validation of the robotic fleet, and the establishment of standards for adaptive and data-driven digital twin integration and mine planning that will serve as a unified tool for all the operational categories and financial decision making.