There is a renaissance of interest in using robotics in mining in South Africa,” affirms Council for Scientific and Industrial Research (CSIR) Materials Science and Manufacturing Operational Unit Mechatronics and Micro Manufacturing (MMM) Competency area research group Leader Peter Bosscha. “We’re ramping up again.” To date, active interest in mining automation and robotics in South Africa has been cyclical: up, down, and now beginning another upswing. (Passive interest never went away.)
The experts within CSIR MMM include electronic and electrical engineers, mechanical engineers, computer scientists, computer engineers and computer vision specialists. “Computer vision is very important,” he notes.
The return of active interest in mining robotics has two main drivers – one predominantly from the industry and the other predominantly from government and public institutions. On the part of the industry, it is rooted in alarm at the future of hard-rock mining in the country. The public sector shares this alarm, but also desires to expand and diversify the country’s mining equipment manufacturing sector. And both sectors are concerned with increasing safety in the mines. So, on the one hand, some mining companies are expressing interest in funding mining automation and robotics research and development (R&D) projects, and, on the other, there is government’s Mining Phakisa initiative, part of its Operation Phakisa (‘Hurry Up’) programme.
“Operation Phakisa is a fast-results delivery programme,” says its website. “Operation Phakisa is an innovative and pioneering approach to translate detailed plans into concrete results through dedicated delivery and collaboration.” A key element of the Phakisa initiatives is developing collaborations between the public and private sectors and academia and civil society organisations.
Research and Restraints
The major outcome so far from the Mining Phakisa is the revival of the CSIR’s facilities located in Carlow road, in Johannesburg, to form what is now known as the Mining Precinct. This is a collaborative effort encompassing the CSIR, industry associations (the Chamber of Mines, the Mining Equipment Manufacturers of South Africa) and national government departments (the Department of Science and Technology and the Department of Trade and Industry). The Mining Precinct is currently implementing the South African Mining, Extraction, Research, Development and Innovation strategy (better known as Samerdi, for short).
Currently, under Samerdi, there are six identified research streams for the Mining Precinct. These are: Advanced Orebody Knowledge (looking inside the rock, ahead of the mining operations); Nonexplosive Rock Breaking (seeking to implement continuous mining in hard-rock mines); Longevity of Current Mines (aimed at extending the lives of current mines by optimising their operations and developing and adopting best practices); Mechanised Drill and Blast (MDB – including improved safety as well as improved productivity); the Real Time Information Management System (improved underground communications and information flow, including basic research into the best frequencies for this, as well as developing an Internet of Things for underground use); and, the Successful Application of the Technology Map (that is, trying to ensure that the technological developments and innovations produced by the Mining Precinct are actually adopted by the miners).
CSIR researchers are involved in all these streams, as are university researchers and industry experts. The first year, which ends in March, has been concerned with research to identify what each of these streams requires to be implemented and the technology or process gaps that have to be filled. The second year, starting in April, will be centred on the development of technology prototypes.
The CSIR MMM Competency Area (based at the CSIR’s main, Pretoria, campus) is focused on the MDB research stream but it also involves other streams. It expects to be involved in creating mining robot prototypes during Year 2. “We don’t know yet what those prototypes may be, as we are still waiting for the Year 1 reports,” notes Bosscha. “Successful prototypes, hopefully, will lead to the manufacture of production units by Mining Equipment Manufacturers Association member companies.”
“One of the things we struggle with a lot is that there is still very little research funding available,” he stresses. “Some people think that the CSIR is fully funded; we aren’t. We need to get funding from somewhere. South African mining companies, [and] South African industry, in general, have been reluctant to invest in research. And small companies do not have the ability to provide funding. Big or small, they all prefer to acquire finished products from overseas. Major overseas mining equipment companies, like Sandvik and Atlas Copco, have significant research programmes, backed by their home governments. With imported technology, maintenance, repair and other support can be very expensive. Why not invest in local technologies? The problem is we don’t have ready solutions – nothing we can point to and say: we can sell you that, now.”
Fortunately, local research funding, from government has been increased as part of Operation Phakisa. Nevertheless, the preference for off-the-shelf products by local miners means that the CSIR (and other local R&D teams) find that they have to develop prototypes that are as close to the production standard as possible, without actually being fully compliant with the production standard – the CSIR is allowed to prototype, not fabricate, production units. “So, while we can develop prototypes to high technology readiness levels (TRLs), we need some company to take it to production standard and then manufacture it.” The CSIR normally takes technology to TRL 5 or TRL 6, while full commercial production takes place at TRL 9.
Back to the Future
Before the Mining Phakisa, the Mining Precinct and Samerdi, there was the Centre for Mining Innovation (CMI). That was closed down, for financial reasons. “But before it was closed, we had a mining safety robot project with the CMI and the CSIR Mobile Intelligent Autonomous Systems Laboratory,” points out Bosscha. “We saw it as an opportunity to showcase CSIR capabilities. We focused on a safety application for robotics because of fears in the trade unions that robots would cost jobs.
“I think that the safety robot is still a good approach,” he affirms. “The prototype, a capability demonstrator, still exists. But it was not and is not really mine worthy. We built an artificial mining stope in Johannesburg to demonstrate that the robot could surmount or avoid obstacles, that it had the traction to move around the stope and that it could carry out inspections – specifically, of the ‘hanging wall’ [roof].”
Fall-of-rock is the biggest cause of fatalities in South African mines. The robot was meant to identify loose, hanging rocks, after the rock face in the stope had been blasted, which would pose a danger to the miners coming back into the stope to remove the ore freed by the blasting. This was possible because loose rocks are slightly cooler than the solid rock around them: the temperature difference, although small, is enough to be detected by thermal cameras. Also, sounds made by loose rocks are slightly different to those made by the solid rock and again the difference is just enough to be detected by an acoustic sensor. So the robot was fitted with a thermal camera and an acoustic sensor and processor. The idea was that the miners would enter the stope only after the robot had carried out its inspection.
“The mining safety robot has been shown a number of times, but is not currently used for active research,” he reports. “But it is in working order. It was displayed at the Mining Indaba.”
Also safety related is the interest, by a number of companies, in a simple robot or automated rig that could load explosives into drill holes in the rock face. Such a system would also increase efficiency by ensuring that all the drill holes were loaded, and properly loaded, with explosives. “We have a few proposals with regard to that,” he reports. “But fixed industrial-type robots do not really belong in mining. Our hearts are in mobile robots – not humanoid robots, but wheeled or tracked robots, which often look like vehicles.”
CSIR MMM is also in the process of importing technologies used in other industries for mining applications. But, with one exception, none of these are applicable to mining. The exception is the SID, which is a CSIR-funded autonomous railway track inspection and safety machine – a railway robot. (SID stands for Survey and Inspection Device.)
The operational concept for the SID is that it would fulfil two missions. Firstly, it would patrol railway lines to detect any buckling of the rails and/or any theft of the clips that hold the rails to the sleepers, as well as to check that level crossings are clear. Secondly, it would proceed ahead of heavy ore trains (iron-ore or coal), which take a long time to stop – their stopping distance is greater than the distance their drivers can see ahead – again to detect any damage to, or failure of, the track, and alert the train drivers in sufficient time for them to safely stop their trains. In all cases, the aim is to avoid accidents, especially derailments. National transport company Transnet has expressed considerable interest in the SID, which exists as a working prototype.
But mining companies also have lots of railway track, of varying gauges, underground. “Some mines have as many as 780 locomotives underground, and derailments are an issue,” points out CSIR MMM Competency Area principal engineer Dr Shaniel Davrajh. In fact, accidents involving rail and road vehicles are the second-highest cause of fatalities underground. “The criteria for inspection on underground mine railways are different from those on conventional railways. However, an underground SID would use the same sensor technology as the surface SID.”
The Way Forward
“There is a lot of local capability in terms of development, but it is not being used,” laments Bosscha. “There are lots of people in the CSIR and universities who, if brought together, could make a real impact.” The key to a successful outcome is collaboration between researchers, robotics manufacturers and mining companies. “Unfortunately, it’s never been that way in South Africa. Mining companies have been reluctant to exchange experiences. But it’s needed for the future.”
“It’s changing – but slowly,” says Davrajh. “Manufacturers working closely with mining companies are also reluctant to share information – probably to protect the mining companies.” It is usually the miners, not the technologists, who resist cooperation. Worldwide, collaboration between State research agencies and businesses to advance mutually beneficial basic robotics R&D is very much the standard model. South Africa has to follow suit.
And there is another key point. “In my opinion, there is no silver bullet that is going to save the industry,” warns Davrajh. “Its going to be a combination of technologies, procedures, training, et cetera. Research indicates that most failures in mines are human failures. It’s not just about the technology, but also about how we use it, where we use it, and when we use it.”