China will Drill the Himalayas! (And Winning the Mining Lottery)

💡What’s New: China will drill the Himalayas! (and winning the mining lottery).

🤔 Opinion: Skipping today b/c there’s lots of news.

🛠️ Tools and Data: A cornucopia of new tools to integrate into your workflow.

Highlight: A timely PhD dissertation on “What Drives Shareholder Returns in Mining?” tldr; scroll to bottom.

💡What’s new

Geopolitics - China is outbidding the US on minerals in Africa. Something you won’t hear most places is that China’s advantage in rare-earth elements has declined..and their drill rigs are rolling into the Himalayas. Last week I mentioned the nationalization of assets in countries like Chile. Subsequent selloffs of foreign assets–like this 14% stake that Vale held in Indonesian mines–is another symptom of a superheated market. Meanwhile, in S. Africa illegal mining is so bad that they’re deploying the military.

Markets & Finance - As mentioned previously, I remain bearish on short-term performance of mineral resources. Commodity cycles typically run long and feature steep peaks and valleys. I am not wholly convinced we’re done with the drawdown that began in ‘22.

Furthermore, mining, like oil, is subject to enormous (and pending) ecological risks. If I learned anything from the Fukushima disaster and subsequent decline in nuclear energy, it is that industries can be put on ice for long periods.

Some interesting news came from BlackRock on how ESG backfired for banks because they didn’t realize mining is critical for decarbonization. Whoops! It means opportunity for VCs and private equity. Big Oil has also stepped in and started mining. I mentioned before that most of the surface-level mineral deposits are assumed to have been claimed. So miners are starting to look deeper, signaling a growing market for underground modeling. Some ideas:

• A list of companies developing tunnel modeling tech.

• Consider old, abandoned mines in the US as opportunities to buy and hold for a billion dollar lottery ticket. There’s an upcoming USGS webinar on the 17th about efforts to identify mine reclamation opportunities.

• Some jurisdictions are promoting mining R&D, and startups should leverage that. Obvious places make sense like Portugal and hi-tech S. Korea.

• In the US, the Biden administration is funding mineral extraction from mine waste as well as mining education.

• Mining is all the rage, but there’s also a need for refining technology. Check out Hackaday’s mining and refining series.

Technology: AI, Earth Observation, Robots - Exploration tech startups seem to be fairing well. Kobold metals has started to drill baby drill! in Greenland. And here’s another AI-driven discovery of critical minerals. If you feel behind in your tech developments, one way to get going is to partner with a tech startup like Comstock did. Or just peruse a list of all the ways AI can be used in mining. You might get some ideas from this project on deep learning to style maps, as well as this one to find places to live in the US that are similar to your town. (Hint: what if you retooled for finding similar mining sites?).

The rush for earth observation analytics is on after Citadel reported netting a $16B+ profit using weather data. Go figure! Hyperspectral imaging was just recognized by Times as one of the greatest inventions of 2023. And soon there’ll be a satellite with dedicated anomaly detection AI in space. One of the sub-themes in earth observation is applying AI to satellite images. Several groups have recently demonstrated how to classify land. (The publication on how it works will help you figure out how to classify minerals on the ground). Before starting a project like this, one must understand the influence of image resolution on the outcome. Several companies provide high-res sat images for a price. That’s why I think it’s worthwhile to mention again the concept of super-res techniques applied to low-res images. These techniques are beginning to proliferate. I haven’t seen a study yet on whether they’re useful based on the noise they generate. But it’s definitely something to watch as it could upset the data provider industry.

Finally, I’m seeing many new robot tech startups. If Amazon can scale from 1k to 750k robots in a decade for warehouse ops, what could happen in mining? More on this development in an upcoming post.

Jobs & Learning - Canadians are rightly concerned about the skills and knowledge gap in mining. Be sure you pick up some alternative skills that next-gen mining professionals need. Start with a look at this list of recommended reading for geospatial experts. If you have a few minutes to spend on education, watch the recent Geo for good summit to learn commodity mapping, geoAI, and Google’s EarthEngine. Then you can move on to a deep dive into deep learning for computer vision with sat photos as well as some research to show you how to deploy it. If you’re attempting to track objects, have a look at how to identify objects in satellite images.

Seeking more ways to apply AI? Here’s a paper on predicting and classifying minerals using computer vision. You can do that with some free satellite photos and QGIS too. Speaking of QGIS, it’s now available to use in the browser with web assembly (more examples here), or in a Jupyter notebook (e.g., to draw geological maps).

I am tracking job opportunities like this job board posts. So email me if you would like personalized updates.

🛠️ Tools and Data

A few resources to speed your projects and make you irreplaceable.

Special mentions - Australia delivers again with a new dataset on magnetic imaging at depth to aid in resource discovery. If you need US wetlands data, it’s now free from Fish and Wildlife.

So what drives shareholder returns?

“We find that commodity prices explained the majority of firm performance annually, but that over the long-run, mineral asset impairments had a much more significant influence on performance. Firms that significantly overperformed the industry experienced minimal impairments while firms that significantly underperformed experienced very large impairments.”

Thanks for reading! Want me to look into a particular topic? Email your suggestions and and I will dig.

David