said. “And we don’t get to shut down every plant for two weeks while we do software updates.”
Delivering GM to the Promised Land of Autonomy
Andrew Farah is using his expertise in embedded systems to reach the ultimate automotive goal.
|Andrew Farah is spearheading GM’s development of autonomous driving technology. (Source: General Motors)|
When Andrew Farah launched his automotive career as a co-op student with General Motors in the late 1970s, he was initially disappointed.
“When I came to the auto industry, they put me in data processing,” he recalls now. “And I said, ‘No, I want to work on cars.’”
In retrospect, however, it now appears GM knew what it was doing. Over more than 30 years, Farah has parlayed a background in embedded systems into key roles on a succession of groundbreaking GM vehicles, including the EV-1, Chevy Volt, Spark EV, Bolt EV and now, its autonomous car program.
“It seems I’ve always been following the latest trends,” he said. His experience may turn out to be critical for GM as it competes with the likes of Waymo (formerly the Google self-driving car project) in the world of autonomous driving. That’s because Farah, unlike so many of the more traditional mechanical engineers in Detroit, has a computer background. He earned a B.S. in computer engineering and an M.S. in electrical science from the University of Michigan-Ann Arbor, then weaved his academic training together with a passion for cars to work on computerized braking systems, touch screens and electrification systems.
Autonomy, however, may be the trend that makes all the others look insignificant, he said. He neatly separates the process of autonomous driving into four key steps – sense, perceive, plan and control. “The four steps are very different,” he told us. “They require different kinds of expertise and they result in different systems on the car.” As chief technical architect at GM, Farah will need to tap into different knowledge bases – from electrical and mechanical to software and even artificial intelligence—to orchestrate his company’s mastery of those systems.
GM tests one of its autonomous fleet vehicles in Michigan. (Source: General Motors)
He predicts that autonomy will happen in a series of advances, starting with relatively simple driving chores and eventually graduating to operation in more complex environments. “Even in the human world, there’s a spectrum of drivers who are comfortable doing some things but not others,” he told us. “Autonomous vehicles will be similar in that way – there will be a spectrum. The point is, we won’t deploy them in environments and situations where we don’t believe they’ll be successful.”
That, however, doesn’t mean that GM is shy about pushing the envelope on autonomous technology. Through its investment in Cruise Automation , the giant automaker is already running autonomous cars (with human drivers on board) on busy San Francisco streets. The goal of engineers is to expose the technology to new situations, and thereby chip away at the number of events where humans might otherwise be called upon to take over the driving