Development of Interactive Dashboards and Intelligent Data Analytics for Visual Decision-Making in The Underground Mining Environment: The Sterkfontein Cave Case Study

In the 21st century, the advancements of computing, Internet of Things and wireless communications are enabling mining companies to collect a vast amount of datasets during risk management. Multipurpose sensors installed at many locations in underground mining environment provide data about the safety, security and mine conditions in near realtime. However, it is more important to convert this data into strategic intelligence, which means the communication of the right information to the right people at the right time to enable them to make optimal and informed decisions on risk control and process optimization. The Sterkfontein caves is a world heritage site, and over the past few years, some deformation was recorded inside the caves. Possible reasons include urbanization in the surrounding area, illegal dolomite mining in the past, weathering, man-made archaeological activities and continuous pouring of polluted and acidic water into the caves system. Acid water has high potential to dissolve the sensitive dolomite, create new and enlarge existing rock cavities, which can destabilize rock formations. The Wits Mining Institute (WMI) has installed a geotechnical and geo-environmental wireless sensor network inside the caves for monitoring and to gain a better understanding of deformation in a cave system. In this research, data analytics and visualization concepts have been applied to the data collected from the sensors in order to identify the trends or patterns inside the caves. Several dashboards have been developed to analyze and visualize the ground movement, wind velocity, temperature, humidity and toxic gases (carbon monoxide) using Tableau code. The data from ground movement monitoring sensors placed at different locations can be analyzed and visualized in relation to each other and in different time steps (minutes to hours, daily, weekly, monthly and/or yearly basis). The analytics can also be applied to the aggregated data based on seasonal, time and/or location variation. The results of this study showed that interactive intelligent dashboards are key to interpret underground data insights and to provide the deep understanding required for smart, near real-time decision-making.