brings innovation to the mining industry

0 tops global competition for predicting dangerous seismic events in active coal mines

On March 15, 2016, machine learning team members captured the first two places in the worldwide Advances in Artificial Intelligence and Applications (AAIA’16) Data Mining Challenge. The winning entries provided the most accurate solutions for predicting seismic events which endanger the lives of coal miners and may result in the destruction of highly specialized coal-mining equipment.

Protecting the lives of miners working in underground coal mines is one of the industry’s highest priorities around the world. From providing the best quality equipment (learn more here) to implementing proper safety standards, mining companies are required to measure a variety of environmental factors in mines where their employees work below ground. Unfortunately, even top of the line monitoring systems can fail to predict periods of dangerous seismic movements which, in extreme circumstances, may result in mining disasters. This year’s AAIA Competition addressed this issue and called on data scientists to develop algorithms which could predict periods of intense seismic activity by analyzing seismic energy flow measurements taken within coal mines. Some others suggested that core drilling services, such as diamond drilling, be used to explore greater depths.

“In the last few years, there has been significant progress in machine learning. This allows us to tackle and solve increasingly challenging problems. It is essential that this knowledge and skills find applications outside the IT sphere: in industrial, medical or ecological domains. Here at we enjoy such projects immensely. By providing solutions to real-life, tangible problems, you can clearly see the value added by machine learning. That’s why I’m very happy with yet another victory by our team, and hope that our work will encourage mining companies to invest in and develop solutions based on machine learning,” says Robert Bogucki,’s Chief Science Officer

203 teams from around the world participated in this year’s competition, submitting more than 3,000 solutions for the challenge. The two best solutions came from members of the team. The top entries will be presented at the Federated Conference on Computer Science and Information Systems to be held in September 2016 in Gdańsk, Poland.


Comments are closed.