The fourth industrial revolution: Bringing AI to the mining industry
Artificial Intelligence (AI) and machine learning are seen everywhere in the media these days, and both have been implemented in a staggering 61% of businesses worldwide. This is a huge jump from the minute 38% in 2016. Companies like Google and Facebook have been using machine learning for years to construct their advertisements, and Investment banks have been using these algorithms to dissect through news and make bets on market trends ahead of their competitors. Now, AI and machine learning has finally reached the Mining Industry, but what does it really have to offer this industry?
The most widely-used Al-powered solutions are in predictive analytics, machine learning, and natural language processing - Narrative Science, 2018.
The most widely-used AI-powered solutions are in predictive analytics (25%), machine learning (22%), and natural language processing (14%). These AI tools are also shown to have a significant impact on many parts of business, including business intelligence (90%), finance (87%), compliance risk (55%), product management (68%), marketing/sales (77%), and communications (43%), according to a new report from the Narrative Science and the National Business Research Institute. The real-world achievements are expected to vary widely, of course, depending on the industry it’s utilized in. In the mining industry specifically, the hope is that AI and machine learning will improve efficiency, productivity, and safety.
According to Jane Zavalishina, CEO of Yandex Data Factory, AI can help miners achieve up to 10% in efficiency savings without making any large-scale capital investments, but by simply producing better predictive models. In the production process - melting metal, improving the quality of ores through benefaction - the natural source material can differ in composition, leaving a prediction to be made on its composition in order to understand how to achieve a standardized result. Material science is typically used by a metallurgist to determine such compositions. For ore such as gold and copper, if it is not rich enough during the benefaction process, cyanide is used to enrich it. However, cyanide is extremely expensive, and because it’s exceptionally difficult to determine how much cyanide is needed during this enriching process, it is generally overused. The use of AI in this particular process would allow for a more accurate predictive model and allow for the exact amount of cyanide needed and no more, which in the long run saves money and is less harmful to the environment.
In the future, AI will pretty much be as ubiquitous as electricity. It will be used in all process that have data and, nowadays, we have data everywhere - Jane Zavalishina, CEO of Yandex Data Factory
Of course, miners have their fair share of reasonable skepticism, with concerns in regards to job losses if AI were to take over. AI has the ability to increase safety within the mining industry, but perhaps at the expense of a number of jobs. Taking conscious steps to ensure collisions on road networks in underground mines does not occur, is of top priority in the mining industry, and AI can assist in eliminating these collisions. AI has the ability to learn from images from set cameras and potentially remove the driver from mining equipment altogether. This, of course, is extremely beneficial for safety, and cost savings for the mines themselves. A few equipment manufacturers are currently offering autonomous fleets, and a number of others are in the process of perfecting working prototypes.
Regardless of the good and bad of potential AI in the mining industry, it will still be a long while before we will see them fully implemented in the industry. There are barriers that are yet to be overcome when facing some of the challenges of implementing AI. The systems can often be fussy about the types of data they require to function properly, and they require a significant amount of data to be trained. Data that many mines often fear they do not have enough of, and thus cannot provide. Zavalishina argues, however, that these barriers are more misconceptions than genuine problems. Most companies, she explains, may not have very modern equipment but typically have some form of digital technology incorporated that will store logs. This data is set to increase as the industry continues to digitize. Data collected for around five years is plenty for the AI.
So it seems that AI is still on its way to fully taking over the mining industry, but soon enough it will be unavoidable. As Zavalishina explains, “In the future, AI will pretty much be as ubiquitous as electricity. It will be used in all process that has data and, nowadays, we have data everywhere.”