Google Self-Driving Car Pulled Over for Driving too Slowly

The Google self-driving car caused a traffic jam while traveling 24 mph in a 35 mph zone. Impeding traffic is an offense in California.

Photo Caption: A Google self-driving car was pulled over by a Mountain View, California traffic officer for going 24 mph in a 35 mph zone, causing traffic to back up.

Maybe Google was right: its self-driving cars do drive too cautiously.

A Google self-driving car was pulled over Thursday for driving too slowly. A Mountain View traffic officer stopped the self-driving car for driving 24 mph in a 35 mph zone, causing traffic to back up near Google’s main campus in California.

Impeding traffic is an offense in California under 22400(a) of the California Vehicle Code. Here’s what the rule says: “No person shall drive upon a highway at such a slow speed as to impede or block the normal and reasonable movement of traffic, unless the reduced speed is necessary for safe operation, because of a grade, or in compliance with law.”

Google’s self-driving cars operate under the Neighborhood Electric Vehicle Definition per 385.5 of the California Vehicle Code and can only be operated on roadways with speed limits at or under 35 mph. However, Google has capped the speed of its self-driving cars at 25 mph for safety reasons.

“The officer stopped the car and made contact with the operators to learn more about how the car was choosing speeds along certain roadways and to educate the operators about impeding traffic,” the Mountain View police department says.

Google responded with the following note on its Self-Driving Car Project Google+ page:

“Driving too slowly? Bet humans don’t get pulled over for that too often.

“We’ve capped the speed of our prototype vehicles at 25mph for safety reasons. We want them to feel friendly and approachable, rather than zooming scarily through neighborhood streets. 

“Like this officer, people sometimes flag us down when they want to know more about our project. After 1.2 million miles of autonomous driving (that’s the human equivalent of 90 years of driving experience), we’re proud to say we’ve never been ticketed!”

Google’s self-driving cars have been involved in 14 accidents, none of which were Google’s fault. Eleven of those accidents have been rear-end collisions where the Google self-driving car was hit by a human driver in another car.




About the Author

Steve Crowe · Steve Crowe is managing editor of Robotics Trends. Steve has been writing about technology since 2008. He lives in Belchertown, MA with his wife and daughter.
Contact Steve Crowe: scrowe@ehpub.com  ·  View More by Steve Crowe.




Comments

Totally_Lost · November 14, 2015 · 12:07 am

So Steve, back to the original point ... when does it become Google’s fault that drivers are hitting them because they are going 10mph or more slower than the prevailing traffic?

Google is using a very little car, likely traveling 15mph below prevailing traffic? So how likely is it that some driver is going to be following a larger SUV that rapidly switches lanes to avoid the Google car, and leads the unsuspecting driver into the rear of the Google car with a 15 mph speed difference, and not enough reaction time and stopping distance to avoid the accident?

I really think Google should take the self driving car back to the test track, until they can match prevailing road speeds, or they are very likely to kill/injure someone very badly before this is over.

Totally_Lost · November 13, 2015 · 11:27 pm

Dunno .... back working on the original 2004 DARPA Grand Challenge a lot of us registered teams, enjoyed the kick-off party, worked hard at the problem, and didn’t get close to being able to compete, or even dream of passing the pre-qualifications. A decade later the sensors are a lot better, better processing power is available, and this still isn’t even close to easy. Tesla is certainly a LOT closer, Google is very hard to second guess.

Steve Crowe · November 13, 2015 · 8:37 pm

That’s a great point. Tesla is out in the real world, although its system isn’t completely self-driving. How do you think Google would perform in a real-world environment?

Totally_Lost · November 13, 2015 · 6:05 pm

The real safety data comes when the car operates in the real world, with real world speeds matching prevailing driver speeds for the area ... if the car can minimize creating accidents (including those where it’s not technically at fault) in real world traffic, then we have something to bench mark with human drivers.

Right now it’s operating in a statistical zone with reaction/response times to traffic events that are a very tiny fraction of what real world drivers maintain.  Lane centering at 25 mph is a tiny problem compared to lane centering at 75-80mph for two reasons ... first the time scale is accelerated by roughly a factor of 3-6 and the correction energies are the square of the velocity. This means it’s pretty safe to for the car to veer sharply at 25mph, but at 80mph the control response has to be significantly slower to avoid over correcting. It also means it’s one thing to have the sensors accurate for 25mph, and quite another for the sensors to have the detail and processing power to handle sensors looking farther forward at faster speeds.

The articles about Google’s self-driving cars completely avoid how different this less than 25mph limited driving is, from real world human drivers.

To some extent ... Tesla is really trying to do real world.

Totally_Lost · November 13, 2015 · 4:54 pm

Steve, I was looking more at the google trend (1.2M miles, 25mph, 14 accidents, 11 rear ends) and reflecting on my own driving experience (53 years, 2.6M miles, up to 80mph, 7 accidents, 2 rear ends while stopped) which include first hand experience with drivers significantly slower than traffic setting up, and SIGNIFICANTLY contributing to, “accidents”.

This is not just simple observation, and uncorrelated without statistical basis.

https://en.wikipedia.org/wiki/Solomon_curve

http://www.soc.hawaii.edu/leon/409as010/albano/albano-report1.htm
“Dr James also cites in his book that federal studies have found that drivers going 10 to 15 miles per hour faster have very low accident rates. Those who travel 20 to 25 miles per hour faster are pretty much on par, in terms of car accident rates, when compared to drivers following speed limits. The drivers that cause the most accidents are those travelling 10 or more under the speed limit and those that travel at 30 or more. Looking at these findings, you can say that “speed variance” plays a more important role than cars that are just “speeding.” Speeding becomes dangerous when a driver’s speed does not match with the speed norms set by the drivers around him”

Steve Crowe · November 13, 2015 · 3:58 pm

Totally_Lost,

I think each of the accidents would have to be looked at on a case-by-case basis. Where I agree with you, however, is that when Google announced plans to make its self-driving cars drive more human-like, that seemed like some sort of omission of guilt, to me. Aren’t self-driving cars designed not to drive like humans ... because humans are terrible drivers? I dunno, I always thought that was weird.

Totally_Lost · November 13, 2015 · 1:36 pm

Hey Google .... They say “Google’s self-driving cars have been involved in 14 accidents, none of which were Google’s fault. Eleven of those accidents have been rear-end collisions where the Google self-driving car was hit by a human driver in another car.”

Isn’t that one of the critical results of driving 10mph or more below the speed limit?  Sounds like some citations for causing accidents by impeding traffic are CERTAINLY in order when people are hitting your bumpers from the rear. maybe next accident the officer will get it right, and attribute partial blame for the accident on impeding traffic, if not full blame.


Totally_Lost · November 13, 2015 at 1:36 pm

Hey Google .... They say “Google’s self-driving cars have been involved in 14 accidents, none of which were Google’s fault. Eleven of those accidents have been rear-end collisions where the Google self-driving car was hit by a human driver in another car.”

Isn’t that one of the critical results of driving 10mph or more below the speed limit?  Sounds like some citations for causing accidents by impeding traffic are CERTAINLY in order when people are hitting your bumpers from the rear. maybe next accident the officer will get it right, and attribute partial blame for the accident on impeding traffic, if not full blame.

Steve Crowe · November 13, 2015 at 3:58 pm

Totally_Lost,

I think each of the accidents would have to be looked at on a case-by-case basis. Where I agree with you, however, is that when Google announced plans to make its self-driving cars drive more human-like, that seemed like some sort of omission of guilt, to me. Aren’t self-driving cars designed not to drive like humans ... because humans are terrible drivers? I dunno, I always thought that was weird.

Totally_Lost · November 13, 2015 at 4:54 pm

Steve, I was looking more at the google trend (1.2M miles, 25mph, 14 accidents, 11 rear ends) and reflecting on my own driving experience (53 years, 2.6M miles, up to 80mph, 7 accidents, 2 rear ends while stopped) which include first hand experience with drivers significantly slower than traffic setting up, and SIGNIFICANTLY contributing to, “accidents”.

This is not just simple observation, and uncorrelated without statistical basis.

https://en.wikipedia.org/wiki/Solomon_curve

http://www.soc.hawaii.edu/leon/409as010/albano/albano-report1.htm
“Dr James also cites in his book that federal studies have found that drivers going 10 to 15 miles per hour faster have very low accident rates. Those who travel 20 to 25 miles per hour faster are pretty much on par, in terms of car accident rates, when compared to drivers following speed limits. The drivers that cause the most accidents are those travelling 10 or more under the speed limit and those that travel at 30 or more. Looking at these findings, you can say that “speed variance” plays a more important role than cars that are just “speeding.” Speeding becomes dangerous when a driver’s speed does not match with the speed norms set by the drivers around him”

Totally_Lost · November 13, 2015 at 6:05 pm

The real safety data comes when the car operates in the real world, with real world speeds matching prevailing driver speeds for the area ... if the car can minimize creating accidents (including those where it’s not technically at fault) in real world traffic, then we have something to bench mark with human drivers.

Right now it’s operating in a statistical zone with reaction/response times to traffic events that are a very tiny fraction of what real world drivers maintain.  Lane centering at 25 mph is a tiny problem compared to lane centering at 75-80mph for two reasons ... first the time scale is accelerated by roughly a factor of 3-6 and the correction energies are the square of the velocity. This means it’s pretty safe to for the car to veer sharply at 25mph, but at 80mph the control response has to be significantly slower to avoid over correcting. It also means it’s one thing to have the sensors accurate for 25mph, and quite another for the sensors to have the detail and processing power to handle sensors looking farther forward at faster speeds.

The articles about Google’s self-driving cars completely avoid how different this less than 25mph limited driving is, from real world human drivers.

To some extent ... Tesla is really trying to do real world.

Steve Crowe · November 13, 2015 at 8:37 pm

That’s a great point. Tesla is out in the real world, although its system isn’t completely self-driving. How do you think Google would perform in a real-world environment?

Totally_Lost · November 13, 2015 at 11:27 pm

Dunno .... back working on the original 2004 DARPA Grand Challenge a lot of us registered teams, enjoyed the kick-off party, worked hard at the problem, and didn’t get close to being able to compete, or even dream of passing the pre-qualifications. A decade later the sensors are a lot better, better processing power is available, and this still isn’t even close to easy. Tesla is certainly a LOT closer, Google is very hard to second guess.

Totally_Lost · November 14, 2015 at 12:07 am

So Steve, back to the original point ... when does it become Google’s fault that drivers are hitting them because they are going 10mph or more slower than the prevailing traffic?

Google is using a very little car, likely traveling 15mph below prevailing traffic? So how likely is it that some driver is going to be following a larger SUV that rapidly switches lanes to avoid the Google car, and leads the unsuspecting driver into the rear of the Google car with a 15 mph speed difference, and not enough reaction time and stopping distance to avoid the accident?

I really think Google should take the self driving car back to the test track, until they can match prevailing road speeds, or they are very likely to kill/injure someone very badly before this is over.


Log in to leave a Comment



Editors’ Picks

Meet Jing Xiao: WPI’s New Director of Robotics
In January 2018, Jing Xiao will become the new director of the Robotics...

Disney: Focus on the Robot Experience
The robot experience included in a business strategy is important not only...

Flirtey Wants Drones to Deliver Defibrillators in Nevada
Flirtey and REMSA have partnered to use drones to delivery automated external...

How Many Robots Does it Take to Screw in a Light Bulb?
Watch a Fetch robot with a custom soft robotic gripper use a...