SAP for Mining Blogs
Read and write blog posts showcasing innovative mining solutions; get practical tips to optimize mining operations and ensure sustainable practices with SAP.
cancel
Showing results for 
Search instead for 
Did you mean: 
jennifer_scholze
Contributor
0 Kudos
See part one of 3 of a paper written by Vimal Gaba (vimal.gaba@sap.com), Senior Director- Industry Business Unit, Mining, Metals and Mill Products, SAP Labs India and Indranil Som (indranil.som@sap.com), Value Advisor - Industry Value Engineering, Mining, Metals and Mill Products, SAP India

This part will focus on Machine Learning - one aspect of (Industry 4.0: What’s Next, 2017).  Future parts will focus on IoT, Big Data, 3D printing, Blockchain and a strong Digital Core.



Mining: Top 5 Digital Innovations for next Wave of Productivity (part 1- Machine Learning

Introduction

Despite facing the slowdowns and global commodity cyclicity, India is a fast-growing economy and demand for minerals remains robust in the country. While mining is often criticized as an industry where innovation and technological changes are resisted and implemented very slowly, the industry is looking at adapting international levels of technology with latest innovations, with a far higher focus. Historically mining is considered to be technically conservative and risk averse, often a bit less aligned with the technological innovations that could significantly impact the industry. The mining industry still conducts its business in mostly the same way as it always has. While mining machinery have become larger, equipment are now more sophisticated, but many mining operations today would be very similar to what it was years ago.

The mining industry is going through an intense period of change and the ability to innovate and improve is becoming indispensable. While the boom and recession are a feature of the mining industry, the response of the industry had been primarily around maximizing volume while continuing with inherent efficiencies in operations during boom phase, while cost-reduction initiatives takes the centre stage during the downturn. It is now being believed that true innovation will drive the next wave of productivity gains and financial growth.

Emerging technologies are set to change the way miners operate over the next decade and help them evolve with newer business models. Delivering improved productivity, cost savings, and safety advancements, these technologies could drive economic transformation in the industry in coming years and also help in becoming competitive globally. We discuss few of these technology trends and the specific impact it could bring about in the mining industry in the subsequent sections.

 

Machine Learning or Artificial Intelligence (AI)

Artificial intelligence or AI, as it is commonly called, is an area of computer science that emphasizes the creation of intelligent machines that work and react like humans. Some of the activities computers with artificial intelligence are designed for include speech recognition, visual perception, translation between languages and decision making, which normally require human intelligence. AI trend has become quite common with AI being used to power voice activated personal assistants like Siri of Apple and Alexa of Amazon, self-driving cars or helping doctors in the treatment of their patients. The AI wave is now making a mark in the mining industry as well. Machine learning algorithms are considered the next step for digital mine transformation. Mining companies have also been working to identify and unlock potentially advanced AI use cases, which can ‘Uberize’ the mining industry.

In the prospecting and exploration stage, machine learning can be used to answer the questions ‘where to explore’ and ‘what lies under the ground’. It can help automatically identify rock faces using well logging data and help in classifying the rock and soil classes using remote sensing data. Using satellite imagery, aerial photography, geophysical maps, drone based monitoring, machine learning can help predict mineral prospectivity or the locations of potential ores. Using previous core drill data, soil samples, mine site surveying data, machine learning can predict targets for drilling. Instruments installed onto drill rigs can provide real-time, automated data accelerating timelines for multiple mining stages and decision making intelligence. Tracking system and devices with wireless communication can monitor ecological parameters like ground water, change in temperature, subterranean ventilation to assess the impact of mining activities which can lead to more eco-friendly operations. Machine learning based predictive algorithm can warn operators and maintenance crew of downtime in critical equipment or probability of pressure spike in pump, hours in advance. Another example of machine learning AI application is in assessing of ore fragmentation in underground and open-pit mines, in less than a minute as compared to hours of manual processing by geotechnical engineers.

Mining operations could have severe impact on the environment and the people and other habitat in the neighbouring region to the mine. One of the key concerns of any miner is how to ensure sustainable operations by protecting the environment and rehabilitate the land. Use of remote sensing technologies like satellite imagery can help monitor any environmental changes and predict changes in erosion, wild life habitats, topsoil redistribution and vegetation. Machine learning techniques can also analyse the risk associated with mine sludge deposits and thus help mitigate the environmental risk associated with mining operations.

The application of robotic technology has far reaching potential for the mining industry. Robotic devices powered by artificial intelligence can perform a range of tasks including drilling, blasting, loading, hauling, ore sampling and as well as rescuing trapped miners. Autonomous load haul dump vehicles using robotic technology are already being used by Rio Tinto in its underground diamond mine in Western Australia. At Rio Tinto’s Cape Lambert port, robots are used for iron ore sampling to ensure that the iron ore product meets the required specifications. The robots work in an enclosed area in the iron ore sample station and are surrounded by numerous devices including in-feed and out-feed conveyors, bucket and tray storage racks, ovens, and weigh scales. Automated and tele-operated drilling can ensure mining personnel safety and improve efficiency during surface drilling operations. The technology allows the operator to carry out drilling from a remote location without entering hazardous areas. Real-time video and data communication including all drilling controls and equipment status are displayed continuously on screens at the operator centre. Mining companies are progressively adopting robotics to stay at the forefront of mining technology. With the huge volume of data being generated at mine sites, these machine-learning examples are only the start of significant value to be derived from machine learning AI. BHP Billitons copper gold mines in Canada uses a number of drilling machines that are partially or fully automated. Another example is Newcrest mining in Australia exploiting machine learning for use in autoclaves (type of chemical reactor). Overall AI and Machine learning algorithms are considered today as the next step of digital mine transformation. The deployment has become simpler and could be done using edge and cloud computing.