Local (303) 454-8000
Toll Free (800) ROSEN-911
Se Habla Español
Contact Me For a Free Consultation

Google Car Programming Takes Cyclists Into Account

| (Google+)
Preventing bicycle accidents

Google illustration of how its autonomous cars can recognize a bicyclist’s hand signals. Image courtesy of the Google Self-Driving Car Project Monthly Report, June 2016

Google’s self-driving cars are programmed in various ways to recognize and accommodate bicyclists, Andrew Krok writes for CNET’s ROAD/SHOW.

Like a considerate robot, Google’s self-driving cars are given instructions to slow down or “nudge over” to provide a bicyclist more space to ride when the Google car’s sensors detect that a parallel-parked car’s door is open.

The latest Google monthly self-driving car report (June 2016) is all about bicycles in relationship to the self-driving vehicles. Google writes in the report:

We also aim to give cyclists ample buffer room when we pass, and our cars won’t squeeze by when cyclists take the center of the lane, even if there’s technically enough space. Whether the road is too narrow or they’re making a turn, we respect this indication that cyclists want to claim their lane.

Typical Riding Behaviors

Google’s engineers have taught the autonomous cars to recognize certain typical riding behaviors, to help the cars to be more able to predict where a cyclist is going. This can be a real challenge, because, as Google writes:

Cyclists are fast and agile — sometimes moving as quickly as cars — but that also means that it’s hard for others to anticipate their movements.

Therefore, Google has programmed the software in its self-driving cars to drive conservatively around bicyclists, who are “unique users of the road.” One way the large tech company is accomplishing this is by having several cycling enthusiasts on its engineering team.

Reading Hand Signals

The cars also have the ability to recognize a bicyclist’s hand signals that indicate the rider intends to turn, or “shift over.” Because a bicyclist sometimes signals with his or her hand far in advance of the turn, the Google cars’ software has been designed to remember that rider’s previous signals in order to anticipate upcoming turns.

Google writes about one tricky situation involving two bicyclists at night:

Our car cautiously approached a cyclist that veered into our lane and stopped to avoid another that suddenly turned a corner and rode directly at us against the flow of traffic. Our car was able to adapt to this unusual situation, and avoid a potential collision.

Increasing Bicycle Use

The report notes that in 2014, more than 720 cyclists were killed, and more than 50,000 injured on U.S. roads. Cycling has become more popular, with the number of bicycling trips increasing from 1.7 million in 2001 to 4 million in 2009.

Google notes in the report that there are thousands of minor vehicle accidents taking place daily throughout the United States, the vast majority of which (94%) are due to human error. Google believes that the estimate that 55% percent of such accidents are never reported is too low, referring to an estimate by the National Highway Traffic Safety Administration in a 2015 revision of a 2010 report.

As of June 30, the Google Self-Driving Car Project’s 24 Lexus SUVs and 34 prototype vehicles had driven a total of 1,725,911 miles in autonomous mode. The company has been testing them in Kirkland, Washington; Mountain View, California; Phoenix, Arizona; and Austin, Texas. The project, which began in 2009, also includes 1,158,921 miles driven in manual mode.


My daughter and I first consulted with Dan Rosen after a very serious auto accident. Dan had several phone conferences with me, and Tracie was available whenever I called. We would recommend personal injury attorney Dan Rosen to anyone!
Sally from Denver, Colorado

Law Offices of Daniel R. Rosen

1400 16th Street #400, Denver, CO 80202
Open M-F 9am to 5pm.
(303) 454-8000 | Contact Us

Don't live in Denver?

We also have offices in Englewood, Colorado Springs and Greeley.
Embed this infographic:
Embed this image: