The Next Thirty Years of Advanced Mining Engineering
The mining industry is finally adopting technologies and solutions at the necessary speed to improve the efficiencies and mining practices in the short term. This change is coming from within. At universities, young professors are enabling a new generation of mining professionals to understand the advantages of applying IT to their daily practices.
Most of the cutting edge research is being conducted by professionals whose common focus is mining data that can be used to improve industry practices. Young professors are exploring topics such as advanced use of computers, operational research, process control, expert systems, decision-making codes and fuzzy logic.
Sean Dessureault, assistant professor at the Department of Mining and Geological Engineering at the University of Arizona in Tucson, is one of these leading mining professionals. He specializes in using data mining tools to improve mine engineering and mine management through the application of algorithms.
What Dessureault is doing is taking data from all the different databases already available at mine sites and consolidating it into one.
"There are several common concepts in the many existing databases, such as 'time.' This factor is stored under different formats depending on the database. I am working on making the databases recognize each other's information. By conciliating all these variables, mining professionals will have more accurate and exhaustive prediction tools that can, in turn, help develop models for better management practices," he explains.
He says that having a centralized data warehouse will avoid the loss of knowledge. In the next 10 years, Dessureault says, the industry will lose about 50% of the work force, and knowledge might get lost.
"When a person is replaced, the new professional inherits a system that he or she has to learn how to use or that he or she chooses to change. The new person develops his or her own new process and the company loses knowledge, time and money in the process," he says.
Dessureault emphasizes that if the mining industry hurries up and incorporates the knowledge and experience accumulated by mining professionals, it'll be easier, faster and less expensive to train new people.
"We can completely change our work processes before the turnover of workers [occurs]. Rather than waiting for the new engineers to develop models of the way a mine works, they'll have all the historical data available. The learning curve will be shorter and analysis will be based more on models and algorithms than on intuition," adds Dessureault.
Trends and Relationships in Real- Time
- The Next Thirty Years of Advanced Mining Engineering
- Application of artificial intelligence, such as neural networks and expert systems
- Statistical process control, such as with SAG mills.
- evolutionary computing, such as genetic algorithms.
- Operations research, looking at production simulation.
- Underground coal, power plant and coal burning operations.
When asked how his research is changing current mining engineering practices, Ganguli explains that there is a growing realization that the vast amounts of data that are gathered-albeit it in realtime (such as sensor data from a process/ equipment) or not in real time (such as cost/ton or weekly production)-could be hiding nuggets of information that he aims to uncover.
"The mining industry is more aware now than ever before that mathematical tool, such as artificial intelligence, can be applied to optimize many areas of mining. We have just started to think of a mine holistically, as an entity that includes the mine, the mill and everything in between," he says.
Ganguli specializes in identifying trends and relationships in real-time. He says that an area where his research will help the most is the identification of the parameters that impact a process.
"For example, intuitively, one may expect SAG mill rpm, noise and bearing pressure to impact power consumption. Yet, your model may show that it has no impact. I have been surprised many times at how you can drastically pare down the list of factors that affect a process. This is important because then management need focus only on a few factors."
Ganguli would also like to draw attention to some of the traditional operations research practices that are more relevant than ever before. Simulations that require vast amounts of data can help a mine explore what-if scenarios, as well as the use of artificial intelligence, when combined with statistical tools.
"Ask any summer interns what they did in the mines and most will tell you they did time studies. What happens to these studies? Not much. They look at some 'averages' such as travel time or cutting time. Averages are fine, but variance is the killer. If this data was used in a simulation, it would be a lot more insightful because we would see how the variances impact your production/process."
The practical implications of Ganguli's research can improve the conditions of the industry. For example, a properly designed database with appropriate built-in queries can instantly provide an intelligent snapshot of a mine.
"I used to work at a mine where we spent the first hour of the day simply repackaging the previous day's data into a management report. That is a waste of time. If the mine information spanned the various departments, repackaging would take seconds. How about your most expensive process? What is the operational data (real time and otherwise) telling you? Most of the time, you will find out that the expensive gadgets you bought do not work well or are unnecessary. You may also find that you need to pay attention to certain factors and not others," he concludes.
Perceptions and Communities
From the University of Kentucky, new ideas swirl around Braden Lusk whose research stretches from quantification of blasting impacts on local communities through mine operations to public relations. In a lively presentation he made at a conference in Denver earlier last year, he referred to modifying public relations practices in quarries adjacent to neighbourhoods that differ in demographics in order to reduce the "felt" or "experienced" impact of the blast on demographically different areas.
"My research has done something that hasn't been done in the past; it asks the public what it is that they would like to know about what we do. I feel that the use of confusing units such as decibels and PPV/frequency often frightens residents concerned about the safety of their homes. I found that residents would be more comfortable with a unit like PSI (pounds per square inch) for air blast," he says.
Lusk adds that gathering some information from neighbours before developing public relations policies is as important as utilizing the most current technologies to sustain mining and improve the quality of life of those residing near mining operations.
"My aspiration is that the results of my ongoing work will be able to benefit mining operations. I think that a large component of integrating mining operations into the community includes a two-way communication that yields positive relationships. The realization is that communities cannot survive without mining; however, mining operations cannot survive without the support of the communities in which they operate," he says.
Lusk believes that the mining industry must take steps to ease communication and strive to keep residents informed. "A key element in this task is to first identify what level of information [is needed] and in what form will it be most easily accepted by the community." Three young men, three different areas of research, one objective: that of making of the mining industry a more efficient, productive and safer business sector.
Links and References:
- An Introduction to Data Mining
- Artificial Intelligence
- Department of Mining and Geological Engineering
- Evolutionary Computing
- Expert Systems
- Introduction to Evolutionary Computing
- Neural Networks
- Production Simulation
- Rajive Ganguli
- Recruitment and Retention Challenges in the Mining Industry
- Sean Dessureault
- Statistical Process Control
- Time Series Analysis
- University of Arizona
- University of Alaska Fairbanks
Click here for full list of links:
Braden Lusk is one of the youngest professors
in the field.