AI for Mining — Part 1

Original article was published by Mohan @OPTOSS on Artificial Intelligence on Medium

Geospatial data in the Mining industry

The business of natural resource exploration and extraction is an inherently spatial one. As such, Earth Observation Data (EOD) has long been considered a valuable source of actionable intelligence for all phases of the mining life cycle: from exploration — extraction — closure — post-closure.

And while EOD certainly is used to some extent, two key factors have persistently impeded the widespread adoption of geospatial data, namely:

  1. the prohibitively expensive costs related to the acquisition of high resolution geospatial data (which is necessary to unlock a new cadre of truly actionable insights for mining sites)
  2. the lack of trained human geospatial analysts at mining sites

The solution: fusion of AI + Geospatial data+ on-site sensing

Tackling these challenges requires the implementation of ​a unified monitoring platform that combines a variety of relevant data sources. From remote sensing and positioning technologies, to Earth observation and Earth GNSS data. Combining these technologies into a unified platform will not only create an opportunity to decrease the associated costs of the underlying data (since the data can be used/re-used by multiple users for various use cases), but the integration of AI processing and data-analytics technologies opens the door for intelligent automated analysis (thereby reducing the need for trained human analysts at any given mining site). Developing this unified intelligent platform is precisely the goal of the Goldeneye project. OPT/NET’s involvement in the Goldeneye project will include playing a technological leadership role and supplying AI expertise for the consortium partners, based on our extensive experience in delivering AI solutions for critical data-driven domains.

While there are many exciting applications that come to mind, the scope of the project hones in on several areas where we believe these innovations will enable the optimisation and potentially automation of processes across all phases of the mining lifecycle, namely:

  • quasi-real-time recognition of minerals during extraction
  • continuous monitoring of slopes, gradients, and terrain elevation changes to watch out for landslide risks (in open pit mining) & terrain elevation changes (in underground mining)
  • tracking of mining equipment and utilisation
  • tracking and precise measuring of waste disposal & environmental risks
  • data fusion & processing to support early warning and smart decision making throughout the mining lifecycle

As the transition to a more sustainably conscious world continues to accelerate, albeit one that still relies on the extraction of natural resources to function, we at OPT/NET are especially excited about the opportunities this project uncovers to find more sustainable approaches to mining. Measuring and minimising environmental risks at mining sites, for example, is an important use case that aligns with our own company ethos. Having previously deployed an award-winning AI-driven disaster management platform — TSAR AI — for humanitarian use cases such as flood-detection and disaster mapping, we believe our drive & commitment to proliferating AI for good causes aligns perfectly with this project and its stated goals.