The mining industry is being transformed by technological advances in both standard digitization and artificial intelligence (AI). Digitalization in safety and production allows mines to work more efficiently, saving hours of labor and operating costs. In many of the world’s largest companies, artificial intelligence is currently being utilized and integrated into almost all aspects of the job.
Mining technology is advancing at an exponential rate and proving itself to be irreplaceable in production and operations management. As this technology merges with our daily lives there are a number of common concerns that must be addressed.
Since the introduction of self-driving cars, experts and consumers have questioned the safety of allowing artificial intelligence to replace humans. The consensus seems to have emerged that these technologies can be more accurate at assessing risk than a human being and can even be better at resolving the situation. Despite the benefits, there are a handful of concerns from industry managers and executives about their implementations.
When addressing these concerns, however, we find optimization of productivity as a major benefit of implementing an AI system. It can help by suggesting the most effective ways to manage:
- Geotechnical Mapping
We had the opportunity to interview Jean-Marc Rousseau, the Director of Technology Transfer at the Institute for Data Valorization (IVADO), our newest partner. He gave us an inside look on how AI will progress in the future and how the mining industry can implement these new technologies.
Newtrax: How is AI being regarded within the mining industry?
Jean-Marc: I look at it from this perspective: we are now able to have autonomous vehicles running in cities without a driver. Because of this I certainly can foresee in the future that new mines will be operating completely automatically, with self-driving vehicles. There will certainly be someone in control above ground, but there will fewer miners underground. If a self-driving car can be trusted on the streets of a city, machinery should also be trusted to drive through a mine.
AI is also able to solve many common productivity problems that arise in a mine. For example, there are many hours of work that are lost when a machine breaks down. One thing that AI could do, with enough data, is it could predict when equipment will need maintenance. This will help companies by saving them time and allowing them to fix problems before they happen.
Newtrax: What are some misconceived notions people in the mining industry have about AI?
Jean-Marc: Not everyone understands AI in the same way. Most people have limited knowledge but the most important thing to understand in AI is what we call “deep learning.” Deep learning, simply put, is when the machine learns by example.
An expert in the field can help the AI machines learn by giving them examples of different situations, and then the machine will copy what it was taught. This is how self-driving cars learn how to drive and how Google Translate learns languages.
Deep learning is what will enable mining machines to make real-time decisions. They will learn what to do through simulations and will be able to react appropriately just as an experienced operator would.
Newtrax: What is the ROI that mining companies should keep in mind when considering the implementation of AI technologies for an underground project? How does the AI help make decisions easier?
Jean-Marc: To start, it possible for AI technologies to learn geographical mapping. It can learn where a company is most likely to find mineralization in exploration phases. It is able to analyze a map and if it has enough data, can learn from where experts were successful in previous attempts.
AI can also help companies by saving them time and money in operations. It will look at the data and make quick, effective decisions. This will help companies cut downtime of machinery, keep productivity high, and help mitigate unpredicted events.
Newtrax: Where does the responsibility lie for researching and providing AI technologies in the mining industry, in regards to the private sector and other academic or research institutions?
Jean-Marc: I think the mining industry must get involved. The software is publicly available for anyone to use. There are even AI scientists that have released their own software and algorithms. The only thing that is not public is the data that these AI machines must learn from. This is how the machines will learn how to be effective and make good decisions through the deep learning the data offers.
So educational institutions can help mining companies if it is a partnership. You can’t have one without the other. We are able to come up with algorithms that will help these companies but they mean nothing without having the data to implement.