Programming vehicles to anticipate the unexpected is the most challenging element of autonomous-driving technology.
(TNS) — Google says it has logged 2 million miles of driving on public roads in its self-driving vehicles and continues to do about 25,000 miles of test-drives a week.
The mileage suggests that the development of autonomous vehicles may be progressing rapidly. Engineers are dealing now with the most difficult part of programming the vehicles — how to anticipate the unexpected, Google says.
The Google vehicles, for example, can now detect whether a moving object is a child or a bicycle and prepare for unpredictable movements, according to Reuters. The car's driving system calculates the probability of such movements to determine how it will react, according to Reuters.
"You have to have a deeper understanding of what's on the road and to the side," Dmitri Dolgov, head of self-driving technology at Google, told Reuters. "Every time you drive, it's different."
Those kinds of complex calculations and reactions are the most challenging element of autonomous-driving technology, Dolgov said.
"You get to 90 percent autonomy in 10 percent of the time and then spend 90 percent of your time on the last 10 percent," he told Reuters.
The giant California-based company -- a leader in the race to autonomy -- is focused on making cars fully self-driving, meaning they will likely not have pedals or steering wheels.
©2016 The Dallas Morning News, Distributed by Tribune Content Agency, LLC.