SFU Researcher Finds Low Cost Solution To High Auto Emissions

The Simon Fraser University engineering scientist is working on an onboard diagnostic system designed to turn on a warning light when car emissions become too high, due to instrument failures, aging and other factors.

Saif is wrapping up a three-year study for General Motors satisfied he has "promising" results that would mean upgrading existing software in new cars, rather than producing new hardware at additional cost, as companies try to become fully compliant with new environmental regulations set to take effect in the U.S. and Canada next year.

"There is a flurry of activity among car manufacturers to fully comply with the new legislation, which calls for all new cars built after 1998 to have the intelligence capability to monitor their own 'health,'" explains Saif, who earlier spent a year on sabbatical working in G.M.'s research and development lab in Detroit. "For many years pollution regulations have existed, but those were a lot looser, and as the concerns over the quality of the environment grow, we are likely heading toward a further tightening of such laws.

"With that in mind, it's important for the auto makers to anticipate more stringent requirements in years to come, and stay ahead of the game by constantly looking for better means of engine control and monitoring."

Using engine data from healthy and faulty research cars at the Detroit lab, Saif came up with a model-based technique which uses mathematical algorithms to detect and isolate engine faults. "Cars already have on-board computerized engine control tasks, and we are certain that for the most part, the same data can be used for diagnostic monitoring, allowing us to do additional things without introducing costly new hardware," says Saif, who as a graduate student in the early 1980s worked on the diagnostics of jet engines, with a NASA research grant.

"Adding one or two sensors is not a big issue in the aerospace industry, but when you're looking at millions of cars and having to mass produce something new, the alternative is appealing," he adds.

Saif, whose current study was funded by an NSERC strategic research grant, says it's now up to car manufacturers to decide if they want to pursue the research further.