February 2011


The need for a global robotic response to coal mine rescue

By Sean Dessureault


Sandia’s Gemini-Scout Mine Rescue Robot is equipped to handle any number of obstacles, including rubble piles and flooded rooms, to help rescuers reach trapped miners safely and efficiently | Photo by Randy Montoya, courtesy of Sandia National Laboratories

In the wake of yet another major underground coal mine accident, this time in New Zealand’s Pike River Mine, I ask myself why industry is not turning to robotics when faced with a rescue effort. The answer, as I sadly discovered, is that no robots designed specifically for mine rescue currently exist. The question is, why?

Aversion to automation

I have been involved in various mine automation projects since the 1990s. The original inspiration for both my research and business was to investigate why automation has not taken hold in mining, and what tools or techniques could prepare the industry for the massive changes that automation would bring. The benefits of automation can be seen in modern manufacturing plants where people, information and machines are coordinated through algorithms and highly effective management to produce a quality product through processes that are efficient, flexible, productive and supremely safe.

In the 1990s, I had the privilege of working with many of the key personalities in the Mining Automation Program (MAP). This groundbreaking group was able to automate every single underground mining process for a sub-level cave. It may have been the best funded underground hard rock automation R&D program, but the technology, although sound, never caught on in the mining industry. Many competing programs in Canada, Sweden and Australia also successfully automated various processes, yet none yielded a significant change in the use of automation by mining companies.

It became clear to me that the success of automation in the manufacturing sector was due to its complete understanding, hence control, of the system. Factories have a firm control over their processes when compared to mines, and much of this has been achieved through the implementation of systems and industrial engineering, alongside the basic fundamental input of all systems engineering: information.

However, work processes are composed not only of information, but also of people, organizational structures and, in most industrial processes, machines. As such, manufacturing changed significantly in terms of factory management before robots became a central part of the overall system. In mining, we cannot skip that step. We must gain much firmer control over mining processes in order to improve efficiency and advance to the next stage in modern mining: automation.

Too many lives lost

In August 2007, tragedy struck the Crandall Canyon Mine, an underground coal mine in northwestern Utah where six miners were trapped due to ground failure. Ten days later, three mine rescuers were killed and six others injured, putting an end to the rescue efforts.

Almost three years later, the Upper Big Branch Mine in West Virginia also suffered a tragedy, where 29 miners lost their lives, likely in a methane gas explosion. Mine rescuers risked their lives by repeatedly entering the mine to look for survivors or recover bodies.

More recent tragedies include:

• Between December of 2009 and July of 2010, Turkey suffered coal mine accidents at the Mustafakemalpasa, Dursunbey, Karadon and Kucukdoganca mines, totalling 66 victims. In some cases, rescuers were put at risk and even harmed.

• A methane explosion in May 2010 in Russia’s largest coal mine, Raspadskaya, killed 66 miners and, once again, rescuers were asked to enter a very dangerous environment.

• In August of 2010, a cave-in at the Carola-Agustina Mine in Copiapo, Chile, trapped 33 miners underground. What followed was a global response: the entire mining industry mobilized, offering help or donations to Chile. The international media was present to witness how our globalized industry, in this case under the leadership of the Chileans, is able to cooperatively execute a process of great technical and organizational complexity.

• In November 2010, 29 miners were killed at the Pike River Mine in New Zealand, in what was likely a methane gas explosion. Over the next few days, various robots, none of which were designed for mine rescue, were used, but failed. Unfortunately, a second major explosion, five days after the initial event, brought about the end of the rescue efforts.

While listening to news reports about these accidents, I thought to myself: Why are they risking the lives of mine rescuers? Why not use a robot? A robot can be outfitted with sensors to detect humans in a post-accident environment far better than a human wearing mine rescue gear. But the truth is, no coordinated global robotic system exists that would have facilitated the rescues. And while the barriers to such a system are many, they are not insurmountable.

Following the Upper Big Branch tragedy, I decided to look into the gaping lack of an effective robotic response to coal mine rescue. An investigation of current mine rescue robotics revealed that the only fielded system was a modified bomb-disposal robot that had proven ineffective to date. More recently, several R&D projects in robotics have been initiated. However, these have focused entirely on technology; no large-scale project has been undertaken to design a holistic system that considers the mine rescue process, IT design, organizational framework for deployment, financing and other essential features. This is quite understandable, since such a complex and organizationally intensive project would be too large for a single mining company or even a single country to undertake. Moreover, no company would willingly, or should rightly, focus on rescue rather than on prevention.

A call to action

The need for such a system is there. These accidents are devastating to the families of the victims and are a major embarrassment to our industry as a whole. We have a collective responsibility to do what we can, myself included. I have spent the first 15 years of my career studying, planning, managing and implementing automation/IT projects. However, my work has never focused on mine rescue, and so I assume some responsibility, even guilt, for not being sufficiently proactive in this pursuit — until now.

In addition to keeping their employees safe and out of harm’s way, governments and companies also want to prevent international embarrassment and continued condemnation of coal mining (and mining in general). Therefore, an international consortium is needed in which the major coal mining countries, principally the United States, Australia, Russia, South Africa, India and China (or any country willing to contribute, such as Canada or Turkey), and coal mining companies could all contribute to the development, deployment and maintenance of a global robotic coal mine rescue system. This consortium would need to be independent and eventually self-sustaining. The robotic systems should be designed to allow for their transportation, deployment and use within a day, or even hours, of an incident, from any mine requesting help, even if they are not part of the consortium.

Consortiums start with a vision, are initiated by selling the vision to stakeholders, are built through cooperation, guided by experience and are sustained through a sense of purpose. In this case, that purpose is saving the lives of miners and mine rescuers.

Sean Dessureault is an associate professor at the University of Arizona’s mining and geological engineering department, where he directs the Mining Information Systems and Operations Management (MISOM) Laboratory.  His consulting company, MISOM Consulting Services Inc., designs, builds and services data warehouses, and advises corporate leadership on technology strategy.

Post a comment


PDF Version