With the scoping study completed, funding secured, researchers selected, and a detailed set of objectives and outcomes outlined, the first three years of the longwall automation project being sponsored by Australian Coal Association Research Program (ACARP) is ready to roll.

Arriving at this point has taken over a year, after the Australian Coal Research board decided to channel significant funding into a major – or ‘landmark’ -- project in 2000. Additional CSIRO and Co-operative Research Centre for Mining Technology and Equipment (CMTE) funding lifted ACARP’s kitty from $3.9 million to over $5 million, excluding support from equipment manufacturers and mines.

Activity over the past year, directed by an industry-nominated steering committee, focused on reviewing international work in the area and devising the scope of the project and its goals. The two key science organisations, which jointly developed the now accepted research proposal, are the Brisbane-based CSIRO and CMTE, under research leaders Mick Kelly and Paul Lever, respectively.

Achieving the ‘blue-sky’ of a completely automated longwall is something both researchers acknowledge is beyond the parameters of the current program. In the next three years, the project is designed to advance longwall automation to the point of on-face observation (ie, with human operators still present but observing the longwall running in automated mode).

What was recognised early on during the scoping study was that full automation would depend on developing sensors capable of replacing the input from human senses.

“If you take people off the face then you’re removing the eyes and ears and all the other senses that people have,” Kelly said. “There’s a whole range of issues that people observe and constantly adjust. To sign off on that we would have to sign off on a whole sensor development program which was too expensive and high risk at this stage of the project.”

Identifying issues such as voids developing on the face and AFC blockages are the kinds of exceptions reporting the human element excels in. In fact, the analysis of previous automation attempts showed there had been a lack of focus on exception issues.

“Automation attempts have only worked in ideal conditions. As soon as problems or ‘exceptions’ occur on the face, operators revert to manual operation and the automation technology is disregarded,” Kelly said. “Even if the automation technology does work in good conditions, unless it produces as much coal as manual operation it will not be used.”

This issue underpins a main challenge of longwall automation: as the longwall is the mine’s prime profit centre, keeping production going takes precedence over everything else, barring operator safety.

Discussions with operators both in Australia and other countries highlighted that the industry wanted automation to achieve production consistency and, secondly, to remove people from exposure to dust.

So, while the initial phase of the longwall automation project does not aim to reduce people numbers, what is intended to change significantly are people’s functionality and their relative proximity to high-risk areas of mining. Different skills will be required to run an automated system. In addition, as Lever remarked, there is always a lot of work to do on and around the face.

“If people have more time to look at standards outbye, that will only deliver improvements,” he said.

The starting point for automating a longwall is the need to know exactly where the longwall shearer is in three- dimensional space, something that has eluded earlier research efforts.

Other CSIRO research projects have been investigating inertial navigation techniques in highwall mining guidance and to measure shearer position.

“At the South Bulga longwall mine we have been able to track the position of the shearer within 100mm of its position, in 3D space,” Kelly said. “Once you’ve got that you have a basis for control of the system.”

The information about where the shearer is and has been will be used to keep the face straight as well as provide the basis for controlling chock movement.

While automated shearer initiation already exists on many modern longwall faces, what is missing is a feedback loop, Lever said. In other words, current shearer movement is not a function of what the shearer has just completed. A feedback loop from the shearer will allow chocks to move forward at a differential rate, depending on what the shearer has done.

“At the moment operators are trying to control the longwall without knowing in real-time where they are, where they’ve been, and where they should be going,” Lever said.

Similarly, current memory-cut systems are premised on repeating the last shear, without any information about what has actually happened. The inertial navigation system will also form an integral part of improving horizon control on the face, the ‘mecca’ of longwall automation, as Lever describes it.

Of the 10 main outcomes targeted in this first three years, improved face alignment is a critical short-term objective.

Transferring all the relevant data across the face will call for the development of a high-speed data highway, possibly using fibre optic cables. This information will feed into an information system consisting of an operator station containing a range of operating sequences, probably presented graphically. The location of the operator station and delineation of sequences – there are 14 cutting sequences currently in use in Australian longwalls – is yet to be determined.

The industry can expect a wide range of results from the project. Certain key components will be able to be commercialised as models for manufacture, but the wide range of equipment and operating conditions means there won’t be available a simple black-box plug-in solution/design for automating a longwall. Each Australian longwall mine will however be benchmarked regarding current automation status. They will then be supplied with a ‘roadmap’ guiding them to what is required to increase levels of automation.

Speaking at the ground control conference held at Morgantown, West Virginia, USA, in August, Kelly said: “The ACARP landmark process has provided the underground coal industry with a tremendous opportunity to develop and implement cutting edge technologies in a package that will provide an automation capability for our longwall operations.”