4.3.6 Technology and innovation
For the past few years, Gold Fields was a fast adopter of best practice technology rather than a pioneer of research and development in areas such as technology. The cost of developing and applying cutting edge practices was simply too expensive.
However, recent advances in digitisation, automation and mechanisation make it critical that we develop strategies to implement new technologies and partner with IT and OEMs that are leaders in the field. A number of technology companies are working on software advances in mining, which can be grouped under the ‘Big Data’ heading, where data is captured by various sources, digitised, analysed and finally leveraged for better decision making. This has multiple applications for mines, such as geological mapping, geotechnical design, fleet tracking and operator safety. We believe that such technologies will provide us with the edge to fundamentally change our cost structure and improve safety.
We have appointed a new member to our Executive Committee to oversee our progress in this area and to oversee the development of three-year technology and innovation programmes in each of our regions.
Gold Fields has started embracing digital mining, advanced analytics and new software technologies and during 2015, we started working with a number of technology companies at our operations in Australia and South Africa to implement these technologies:
Mine Vision Systems (MVS) – In 2007, Gold Fields started work with Carnegie Mellon University (CMU) in the US to develop mining robots. While the robots were ultimately unsuccessful, the robotic vision component from this programme went on to the used in military applications, oil and gas, heavy industry, autonomous cars and mining. Robotic vision is one the fast growing and influential technologies on the world today.
In 2015, Gold Fields asked CMU to spin out the technology for mining, with Gold Fields providing the initial US$3 million seed funding to create MVS. This gives Gold Fields a 10% share in MVS with a first right access to technology developed by the company.
Since then MVS has put in place 30 non-disclosure agreements with a number of leading software providers and equipment manufacturers and is in early conversations with another 40 interested companies.
MVS provides a system to collect mining data using cameras, sensors mounted on machinery and Light Detection and Radar technology. (Lidar) This data can then be converted into meta-data, compatible with most major mining software providers today. This data is already being used to identify geology, ore fragmentation, convergence and is the basis for geotechnical modelling and machine automation at our mines in Australia. Robotic vision is one of the fastest growing technology areas in the world today.
GlassTerra – This Australian start-up technology company, staffed by geo-spatial software engineers and mining experts, is tackling the impending big data challenges facing global mining companies as the amount of digital mining data available grows exponentially.
Gold Fields worked with GlassTerra to run the Ore-X Challenge in August 2015. Ore-X was the world’s first open crowd-sourcing challenge to solve operational problems in the mining industry. Glass Terra made Gold Fields’ 3D geo-spatial data available and asked experts from around the world to develop an image algorithm that can classify ore material in a gold mine as high, medium or low-grade, or waste.
Gold Fields offered A$12,500 (US$9,000) in prizes to the top three entries. Almost 270 contestants participated and the winning algorithms have been built into a geology software package, named Leap Frog, which allows mining data to be converted into easy to use software. This programme has been expanded to collect and evaluate data from new MVS sensors at our St Ives mine.
Cyest – This South African technology company is working at our South Deep mine to scientifically determine the capacity of the mine’s full value chain. In addition, the firm is developing an advanced visualisation of South Deep, to convey the complexity of the mining process to internal and external stakeholders. Cyest is using four different systems and solutions at South Deep:
- Advanced simulation of the sequence of activities and equipment interactions along the underground value chain
- Validating and improving the mining schedule by modelling the interaction of different mining activities as a function of mine layout, efficiencies and other factors
- Modelling the link between operational drivers to identify what interventions are needed to achieve the desired financial results
- Using an advanced gaming platform to create a high fidelity visualisation of the ore body and the associated mining methods