Check out ITS Basic Training: Bootcamp for Your Traffic Team for Employers & Academic License for EducatorsLEARN MORE
LEARN MORE

blog

RAND Study Examines Safety of Autonomous Vehicles

Subscribe for Updates

Join the Transportation Tech mailing list today and get a free copy of our white paper submitted to ITS World Congress!

Sign Up
by Aaron Mack in Uncategorized
November 27, 2017

If you learned that tomorrow morning, every car in America would suddenly be replaced by a self-driving vehicle, would you (a) stay up an hour later knowing you could sleep on the way to work, or (b) buy a cache of bottled water and sundries, hunker down at home, and ride out the impending roadway apocalypse?

If you’re like half of Americans, you would probably fall somewhere in-between. Autonomous vehicles are exciting! But will they be they safe?

According to a (2014) study by the Pew Research Center, Americans’ opinions on autonomous vehicles are divided, with just over half stating that they would not even ride in one were they given the chance. Granted, this is 2017, and autonomous vehicles have come a long way in three years. Every day, it seems, there’s a new story in the media about progress made, milestones reached, and the impending self-driving revolution. Waymo, for example, announced this month that it will launch a self-driving ridesharing service this year that will soon be open to the public.

But is it all hype, or will autonomous vehicles flood our roadways sooner than we may have thought? And if self-driving car companies say they are ready to put their vehicles on the road soon, should we be excited? Nervous? Or a little bit of both?

This is a rational question, particularly for a computer-based society that knows the performance difference between a product beta test and software update 19.2. We live in a world captivated by new technologies, yet we all know not to expect flawless performance from a first-generation model. That’s all fine and good when we are talking about, say, the latest smartphone operating system. When it comes to our vehicles, however—that is, when it comes to our safety, and that of our loved ones—it’s suffice to say we should be much more demanding when it comes to performance.

Still…given that autonomous vehicles are anticipated to bring some tremendous safety benefits, i.e., accidents avoided and lives saved, it makes sense that we should want them on the road as soon as possible, even if they aren’t yet “perfect.” Right? But where do we draw the line? How ready is ready? How safe do autonomous vehicles truly need to be (or how safe do we need to perceive them to be) before we should be willing to share the road?

A study recently released by the RAND Research Corporation examined this very question. Using a research method called robust decision-making, the study explored the possible safety benefits and drawbacks of deploying autonomous vehicles when they are (a) just 10% safer (in terms of performance) than the average human driver, or (b) 75% or 90% safer.

The researchers found that, even when just 10% safer than the average driver, autonomous vehicles would still probably result in tens of thousands of additional lives saved over just the next 10 years. In the long term, that number could be as high as half-a-million. In other words, if every car in America began driving itself tomorrow, there would inevitably be accidents–but countless more lives would be lost if society were to wait until autonomous vehicles performed flawlessly, or almost flawlessly, before allowing them on the road.

We may not know until autonomous vehicles become more commonplace how “ready” they truly are, but it seems that waiting for perfection may be the bigger gamble.

 

 

 

1 Comment
  1. James R Eidson says:

    What if vehicles were simply selectively “autonomous” based off the HEALTH of the driver? IF a driver were wearing a fitbit, or galaxy watch and had a incorporated heart monitor that were tethered to the vehicle via bluetooth and they had a medical emergency example “heart attack” the vehicle could sense that, ask the driver “are you OK?” and then assume control of the vehicle pull it over, alert law enforcement, and give a google map location pin. The vehicle could flash all lights half cycle from a standard flash pattern and alert fellow vehicles of the issue via nearby ITS signs in hopes of Medical Professionals in the area providing temporary assistance. If you present this idea to ANY individual and tell them an autonomous vehicle could save their loved ones life and then ask them their feelings regarding selective autonomous vehicles they would most likely agree in any way or form.

Leave a Reply

Your email address will not be published. Required fields are marked *

Subscribe for Updates

Join the Transportation Tech mailing list today and get a free copy of our white paper submitted to ITS World Congress!

Sign Up