Geospatial AI for Continuous Multi-Scale Risk Monitoring of Tailings Storage Facilities

Tailings Storage Facilities (TSFs) represent one of the most critical and high-risk infrastructures in the mining sector, with failures leading to severe environmental, social, and economic consequences at local and transboundary scales. Increasing climate variability, ageing facilities, rising demand for mined products, and rising regulatory expectations necessitate more advanced TSF monitoring approaches.  Existing TSF monitoring is often fragmented, as Earth observation, in-situ sensing, and risk assessment tools operate independently, limiting their effectiveness for continuous risk assessment. This underscores  the need for integrated, multi-sensor monitoring approaches that can provide continuous, comprehensive, and predictive assessment of TSF stability and associated risks.
The GAIA-TSF (Geospatial Artificial Intelligence Analysis for Tailings Storage Facilities) project, led by an international consortium, aims to design and develop a prototype system. This system integrates satellite Earth Observation (EO) and ground-based sensor data with machine-learning (ML) algorithms to enable continuous, multi-level, and multi-scale characterization and monitoring of TSFs.
As a work in progress, the project has undertaken a comprehensive stakeholder engagement process to identify current gaps, operational needs, and priority monitoring requirements for TSFs. A review of the state of the art in available EO and ground-based monitoring technologies has been conducted, leading to the identification of key technologies and ML techniques. An extensive review of the literature, coupled with stakeholder input, led to the identification of key variables relevant to TSF monitoring. Such parameters include water quality, air quality, and slope stability. In parallel, potential test sites across different continents have been selected to support future calibration and validation of the prototype under diverse geographical and climatic conditions. The functional requirements and system architecture have been defined, identifying the key components of the prototype and how they are connected. The initial development phase of the GAIA-TSF prototype has commenced.
Integrated TSF monitoring supports risk-informed life-cycle management of TSF, enabling loss prevention and effective asset stewardship. It also strengthens decision-making for ESG compliance, the Global Industry Standard on Tailings Management (GISTM), and climate adaptation, ensuring safer and more sustainable mining operations.
The GAIA-TSF prototype offers a transferable and scalable continuous monitoring solution that enhances early anomaly detection and supports risk-informed decision-making. It thereby contributes to more sustainable and resilient TSF management.