How Harnessing Computer Vision and Machine Learning Will Revolutionize Global Mining

"In the global mining industry, there are those who seem resigned to the fact that the mining industry has always operated according to certain time-tested principles that don’t involve jumping on every new technological trend. Others expect that mines will soon have autonomous electric vehicles running around the clock. The truth lies somewhere in the middle. In business, as in life, that is usually the best path to follow.The instant access to people and information provided by 21st-century technology continues to push businesses in all industries toward greater efficiency and unheard-of levels of modernization. And in the world of mining, the embrace of advanced technology is all about managing risk.However, even in this age of digital information overload, some mine sites still require employees to physically download data from individual trucks to monitor payload data and inspect road conditions using a clipboard. What’s more baffling is that other mine sites still have no way of tracking tons moved at all.This isn’t due to a lack of innovation in the industry. Quite to the contrary, there is an abundance of new technology available that provides real-time monitoring of vehicle movements and tracks road conditions without the intervention of people. So why are owners reluctant to embrace advances in mining tech?Gauging the value of changeSurvey data has traditionally been captured by ground laser scans and LIDAR systems. A mine site surveyor would drive around the pit with a truck-mounted scanner to generate a point cloud. This process would be completed every few days and would take an entire day.The many cons of this method far outweigh the pros. First, it provides no visual data. If an operation has moved beyond the roller wheel and tape measure, visual pit models and LIDAR scans are typically completed once every six to 12 months by an aerial flyover, but such sporadic information gathering isn’t likely to result in the most actionable data. Moreover, given the high cost (in terms of both money and time), the numerous safety and liability issues, and the lack of flexibility in scheduling, manual processes are more trouble than they are worth.Overall, while miners are excellent operators who value innovation, their employers often try to outsource it or acquire innovators instead of investing in existing technology that can improve current operations. It is a never-ending dilemma:"