Oil Sands: New centre for intelligent mining systems
In a new laboratory in the University of Alberta’s computing science department, mouse clicks and streaming video feeds are providing answers that might one day help as much as heavy equipment and hardhats in extracting oil from bitumen in Alberta’s oil sands.
Although it only officially opened in May, the Centre for Intelligent Mining Systems–a “collaboratium” between the university’s computing science faculty and Syncrude Canada Ltd.–has worked on solving production problems for the sprawling Fort McMurray-based mine for the past 18 months.
“It is applying technologies in the areas of artificial intelligence and robotics to the problems that we find in surface mining,” says Ron Kube, a computer scientist with Syncrude’s research department. Syncrude paid $500,000 to lease the lab for three years and outfit it with new equipment.
“Mining isn’t just done by engineers; it’s also done by computer geeks,” says Randy Goebel, the university’s computing science department chair. “Syncrude provides the problems and problem-specific knowledge and we provide the ideas about how to solve the problems.”
The centre has already made major progress in solving an expensive problem. Syncrude mines the oil sands with massive hydraulic loaders, which cleave the bitumen out of the earth with shovels with serrated edges. Occasionally, a steel tooth, weighing between 650 and 800 pounds, will break off a blade and be buried in the 420-ton loads of bitumen. When the broken tooth reaches the crusher, which grinds the clumps of oil sands in order to be processed into oil, it will jam or break the machine. The repair bill combined with loss of production during down-time costs “millions of dollars,” Goebel says.
To overcome this problem, university researchers have worked on an automated computer vision project that monitors the toothline of the mining shovels and alerts the operator when a tooth breaks off. “Computer vision can be tuned and trained to be a lot faster than ordinary people,” Goebel says.
The centre’s success has already prompted talk of a long-term funding proposal worth $1 million annually from a consortium that includes Syncrude. It has also attracted attention from other industry giants such as Noranda Inc. and Inco Ltd., developments that gratify Goebel.
“Science doesn’t stop at the university boundaries,” he says. “The bottom line is you have to be connected enough with the world to understand it well enough to have an impact in it. The ivory tower stuff has never really worked for us.”
Will Gibson is a freelance writer based in Morinville, Alta.
These photos show a tooth that has fallen off a loader shovel, and stuck in the rollers that crush the oil sands ore. The tooth has damaged the crusher. Recent research at the Centre for Intelligent Mining Systems has developed an automated computer vision system to detect when a tooth is missing from a shovel, so it can be retrieved before damaging the crusher.