Goldcorp and IBM Watson: Gaining New Exploration Insights Through Artificial Intelligence

Mar 3, 2017 6:30 PM ET

As part of our series on #DisruptMining Innovation, in today’s post we’ll share how we’ve been working with IBM Watson to potentially transform every facet of the mining process.

Value creation in mining has shifted from how well a company moves material, to how well it acts on information to increase efficiency and production. This shift is highly influenced by the vast amount of data mining companies are able to collect from sensors, equipment monitors and other digital and connected devices. The key to unlocking this value lies in harnessing the capabilities of cognitive computing systems to turn big data into useful insights for future events. 

Cognitive computing, also referred to as Artificial Intelligence, is now considered an essential tool in helping companies make better business decisions on how to improve processes and outcomes. This is one of the main reasons Goldcorp recently started working with IBM Watson.

Unlike traditional computer systems, which are programmed by people to perform certain tasks, cognitive computing systems have the ability to learn through their interactions with both data and humans. Watson, for example, is able to distinguish between different types of information. It gathers relevant data from different sources, creates hypotheses and continuously checks to identify the most reliable and accurate results.

IBM Watson has a unique ability to read, analyze, and learn from natural language, allows it to make informed, context-specific decisions one would only expect from a person, as opposed to a conventional programmable computer.

“At a certain level of complexity, a programmable system will reach its limitations,” notes Dariusz Piotrowski, Global Leader of Cognitive and Analytics Development for Natural Resources at IBM. “A cognitive computing system, by comparison, will learn and improve over time through interaction with experts and feedback from users, essentially reprogramming itself.” 

Many companies are using cognitive computing to capture the expertise of their Engineers and Geologists as a way of retaining knowledge after these professionals retire.

Watson has been successfully implemented in a variety of industries ranging from medicine to oil and gas. However, Piotrowski says the most promising application of Watson in the mining sector is in mineral exploration. This is due to the system’s ability to learn about geology and successful exploration techniques in order to guide Geologists to more prospective areas. Watch IBM’s video on how Watson is learning to think like a geologist:

“We’ve spend a lot of time with the exploration group at Goldcorp,” says Mark Fawcett, Partner, Global Business Services at IBM. “One of the biggest challenges facing Geologists is interpreting the vast amount of data which includes: field mapping data, geochemical surveys, drill hole data, geophysical surveys, geological maps, Landsat imagery, aerial photographs, mine level plans, alteration models, resource model data and reports in order to make the best decision on where to drill next. Inputting that information into Watson and educating Watson on the process of exploration, gives the system the capacity to make more informed exploration decisions to improve the probability of discovery.”

IBM started working with Goldcorp’s exploration group last summer to determine how cognitive computing could improve the company’s mineral exploration program, starting with the Red Lake District.

The first phase of the program (which is still in process) involves providing Watson with structured data from a wide range of sources, such as geophysical and geological surveys, drillhole datasets, reports, academic papers and conference proceedings, to providing insights into Red Lake’s geology, current and historic mining and exploration activities and successful exploration techniques.

“Phase two is where we start to educate Watson to refine and build the exploration model the system will use to make predictions and provide recommendations on where Goldcorp should direct its exploration activities,” says Piotrowski. In an aim to provide a wider perspective on exploration techniques, Goldcorp plans to work with leading universities to access top talent in earth sciences in educating Watson.

“I’m extremely excited about the work we’re doing with Watson in the Red Lake district,” says Todd White, Chief Operating Officer of Goldcorp, noting that exploration is just the beginning. He further stated, “The ability to gain new insights from cognitive systems like Watson has the potential to transform every facet of the mining process. It’s a big part of the future of mining.”