Sean Dessureault and his students and colleagues at the University of Arizona, Tucson are advancing the state-of-the-art in mining mine data. Neither of their two most recent publications are available electronically. You will have to contact them directly to get a copy if you do not belong to both the CIM and the SME. Contrary to my normal policy, I write about these papers, simply because they interest me so much, and I suspect will be important to you and the future planning and management of mining operations.

In the August 2007 issue of the SME’s Mining Engineering is a paper Data Mining mine safety data by S. Dessureault, A. Sinuhaji, and P. Coleman. They describe how they compiled a modern data mining system to go back through MSHA accident records collected since the 1980s. Much has changed in the practice of computer recording, storing, and sorting of data since the 1980s. Basically they reconfigured the data, reconstructed the records, and applied decision making theory to establish dependency networks for injury prediction.

In the CIM Magazine for September/October 2007 (and on the CIM website) is a paper Data mining, mining data: energy consumption modeling by S. Dessureault. To give you a slightly wider view of what he is doing, here is a bit from the executive summary:

As additional efforts and technology are directed to developing even more information sources, a new technological focus should emerge: how to concentrate data into information; analyze information sufficiently to create knowledge; and finally, to act on that knowledge. Stated simply thus:

data => information => knowledge => action

A third paper by S. Dessureault, M. Yildrim, and M. Baker, Data mining mine data: truck-shovel fleet management systems, is summarized in the CIM Magazine of June/July 2007. To quote: this paper describes the application of modern data miing techniques on truck dispatching systems in a real open pit hard-rock copper mine. The long-term benefits of this work are to identify strengths and improvement opportunities in truck assignment algorithms and to establish the skill sets and information technology infrastructure needed to undertake more complex data-driven technology, such as a truck dispatcher trainer.

The detailed procedures are too complicated to summarize here; but recommended if your company has lots of old data and does not know how to sort through it for patterns, connections, causality, or formulation of rationale management actions.