India is home to abundant natural mineral resource and is one of the top ten producers of several minerals (Fig. 1). With a contribution of 1.53 per cent of gross domestic product in 2017-18, mining is an important sector for the Indian economy (Dash, 2019). There is growing realisation that the mining industry can significantly bolster growth in India over the next decade as it will directly impact a wide-array of industries including automobile, cement, etc., impacting crucial infrastructure needs, such as development of road networks among others. As per a 2014 report of McKinsey, the mining industry can contribute USD 47 billion to India’s GDP by 2025 (Mckinsey &
However, mining is plagued with multiple challenges. For instance, to produce a kg of aluminium, 5 kg of bauxite is needed along with 13 l of water. The extraction process itself will demand 15.7 Kwh of electrical energy. Thus mining aluminium not only requires bauxite but also makes significant demands on other resources. Often such demands take a heavy toll on the environment as well. To achieve the full potential of mining it is thus important that regulatory policies, available technologies and human capital work in tandem. The mining industry is usually characterised by several phases:
Exploration—Identifying precise geographical locations where there are significant ore concentrations;
Development—Building infrastructure to aid in the extraction of minerals;
Extraction—Recovering raw minerals, processing and transporting them; and
Closure and Reclamation—Minimising adverse impacts on environment to ensure that the land returns to its original state once the mine is
closed, when the mineral reserves are substan-tially depleted.
The importance of improving the efficiency involved in the various stages of mining cannot be overstated and companies world over are turning to artificial intelligence (AI) to solve these vexing problems. The transformative potential of AI in delivering technological solutions to complex industrial and societal problems is spurring governments around the globe to formulate national policies on its usage. This article illustratively outlines how AI can help address mining concerns in exploration and
AI for Mineral Exploration
The most critical stage in mining is to identify places which has significant exploitable mineral reserves. It is estimated that India may have large reserves of resources that is yet to be discovered with accounts claiming that the volume of remaining reserves could perhaps be twice that of the current estimates (FICCI, 2013) (Fig. 2). Identifying these reserves would require significant investments in technology and several reports have identified this to be a key for improved efficiency of mineral exploration in India (FICCI, 2013; FICCI, 2018; Mckinsey, 2014). The potential of AI in improving the process of mineral exploration is huge and the point is illustrated through a couple of examples.
Finding gold reserves and Kriging: Data analysis is the cornerstone of AI and has a long history in aiding the identification of mineral reserves. Geostatistics, the application of mathematical statistics to spatiotemporal datasets in various branches of geology has played a stellar role in mineral exploration. The first such application of geostatistics goes back to the 1950’s when Danie Krige invented a technique called Kriging, more commonly known as Gaussian process (Krige, 1951) and used it to accurately predict the value of gold reserves in a nearby mine. Since its introduction it has been successfully applied to mineral exploration and still remains a tool of choice. In recent years the ability to collect and process data from a single drill-hole easily exceeds hundreds of mega-bytes. Mining in an area will involve several such drill-holes and analysis of associated data will require tools for data analysis developed in the broad field of AI. Gold Spot Discoveries Inc., was in fact able to predict 86 per cent of the existing gold deposits in the Abitibi gold belt region of Canada by fusing heterogeneous data-sources including geological, topography, and mineralogy from just 4 per cent of total surface area. This is a significant development which demonstrates the promise of AI in mineral exploration (Holmes, 2019).
Ore fragment assessment: Usually ore fragment assessment, an important aspect of mining, is conducted manually. A data science company, PETRA developed an AI algorithm for ore fragment assessment which is fully automated (Petra, undated). Globally there are ongoing efforts in leveraging such data analysis techniques for mineral exploration.
Autonomous systems to improve
Apart from mineral exploration, AI can also help in impacting the various processes involved in mining. Robots, drones, unmanned ground vehicles (UGVs) are examples of autonomous systems which can play a significant role in mines. An Australian mining company Rio Tinto, announced at the beginning of 2019 the introduction of Auto Haul, a fully autonomous train that will help transport iron between various ports owned by the company. The project uses about 200 locomotives over 1,700 km of track to transport ore from 16 mines to four port terminals in the Pilbara region in Australia. Rio-Tinto is in the process of completely automating their processes which would include autonomous loaders that excavate dirt and autonomous blast-hole drillers.
Also, Volvo announced in 2018 that autonomous trucks will be used for transporting limestone from a mine in Norway to nearby ports (Sawers, 2018). Trucks operating on the surface can access the Global Positioning Systems (GPS) which can be used to guide such autonomous vehicles. However, the underground operation of such trucks remains a technological challenge.
Recently an Indian company, ATI Motors, have made remarkable strides in developing a cargo vehicle which can navigate autonomously without GPS to provide logistic support in challenging environments such as mining. The driverless vehicle has been built from scratch and is simple and sturdy. For instance, unlike traditional vehicles retrofitted with autonomy, it does away with the cabin for a driver and hence saves on both space and the ergonomics that goes with it. It uses novel algorithms which can operate on Light Detection and Ranging (LIDAR) and images from camera, combined with inertial measurement units (IMU). It also does not need any augmentation of the external world with beacons etc., to navigate. Unlike traditional automated guided vehicle (AGVs) that operate on fixed routes, the routes on this vehicle can be dynamic which is ideal for mining.
Apart from UGVs, drones are also used in the mining industry. Though in its early days but it is already seen that surveying can be easily done by deploying drones.
Asteroid Mining: Going Beyond Earth
Based on current reserves on the earth and the growing consumption, it is estimated that the raw materials needed for sustaining human civilization would be exhausted within next half a century (Cohen, 2007). It is conjectured that in the not so distant future we will have colonies in outer space. Building such colonies would not be viable if items have to be transported from earth. It is believed that extraction of raw materials from asteroids and other minor planets, could be the key to creating such colonies. It is no longer in the realm of imagination and there are several start-ups trying to design technologies for asteroid mining. Planetary resources, an American Company, plans to create a Fuel Depot in Space in 2020 for refuelling rockets with liquid oxygen and liquid hydrogen obtained by splitting water harvested from asteroids. Though the potential of asteroid mining is enormous, crucial to this endeavour would be the ability to execute the mining process efficiently in space. Development of sophisticated robots suited for these tasks will be thus key to the success of this programme.
It is clear from the illustrations above that the potential of AI in transforming mining industry is huge. Acknowledging the transformative role of AI, governments around the world are now formulating policies on how best to take advantage of technologies arising from the field of AI for betterment of society. The Indian government through NITI-Aayog has come out with a broad strategy plan on how to foster AI to develop technological solutions which can address the needs of the nation. It would be very helpful if all stakeholders in mining industry can come together to devise a similar plan which can specifically leverage AI technologies for more efficient mining.
Cohen D., 2007. Earth’s Natural Wealth: An Audit. New Scientist. Available at: https://www.newscientist.com/article/mg19426051-200-earths-natural-wealth-an-audit/
Dash J., 2019. Mining’s share of India’s GDP fell to 1.53% in FY18 from 1.93% in FY13. Business Standard, June 20. Available at: https://bit.ly/2knwpFb
FICCI, 2013. Development of Indian Mining Industry – The Way Forward. FICCI Mines and Metals Division. Federation of Indian Chamber of Commerce and Industry Federation House, New Delhi. Available at: https://bit.ly/2jW1Fus
FICCI, 2018. Indian Mining Industry A Different Perspective, Report. Available at: https://bit.ly/2k3p6lO
Holmes F., 2019. This AI Company is the Future of Gold Exploration. Forbes. Available at: https://bit.ly/2lxosNE
Indian Bureau of Mines, 2018. Indian Mineral Year Book. Indian Bureau of Mines, Civil Lines, Nagpur. Available at: https://bit.ly/2lw8K5s
Krige, D.G. (1951) A Statistical Approaches to Some Basic Mine Valuation Problems on the Witwatersrand. Journal of the Chemical, Metallurgical and Mining Society of South Africa, 52: 119-139.
Mckinsey & Company, 2014. Putting India on the Growth Path: Unlocking the Mining Potential, Report, December. Available at: https://mck.co/2lYFswB
Petra, undated. Machine Learning AI Enters Underground Mines. Petra Data Science Pty Ltd, Toowong Tower, Toowong, Qld, Australia. Available at: https://bit.ly/2lRwVLJ. Accessed on September 3, 2019.
Sawers P., 2018. Volvo’s First Commercial Self-driving Trucks will be Used in Mining. Venture Beat, November 20. Availableat: https://bit.ly/2ORuGB2