Ready to Take Over? Driver Distraction in the Age of Automation
March 30, 2022 | Webinar
Distracted driving is a dangerous - full stop. Taking your eyes or mind off the road, or your hands off the wheel while driving, can significantly increase your crash risk. But what happens to driver attention when automated vehicles come into play? Will drivers adequately be able to focus on supervising the technology? Dr. Ian Reagan and Dr. Alexandra Mueller joined us from the Insurance Institute for Highway Safety (IIHS), a leading authority on crash risk, to discuss IIHS’ new framework for rating partially automated vehicles and how well they help drivers stay focused on the road.
This program, kicking off Distracted Driving Awareness Month in April 2022, was held as part of the Travelers Institute Every Second Matters® distracted driving education initiative.
This installment of Wednesdays with Woodward® webinar – hosted by guest moderator Jessica Kearney, Assistant Vice Present of the Travelers Institute – opened with remarks from Emily Stein, President of the Safe Roads Alliance, an organization she founded after her father was killed by a distracted driver. “I felt the strong need to talk about what happened in hopes of educating drivers about the power they hold when they get behind the wheel, the power to take away a life in an instant,” she said. Urging manufacturers to champion safety over convenience when designing in-vehicle technology, she added, “No one should be able to order Starbucks from the driver seat. If we do this correctly, some partially automated features can save lives instead of causing more distractions.”
Here are the top takeaways from Ready to Take Over? Driver Distraction in the Age of Automation:
Click each key point to jump directly within the webinar to watch and hear more.
Traffic fatalities increased during the pandemic. According to early projections from the National Highway Traffic Safety Administration, more people were killed in motor vehicle collisions in the first nine months of 2021 than any year since 2006. Stein noted that the pandemic has impacted people’s driving – “it seems that people are driving faster and more aggressively.” She believes fatigue from the pandemic is to blame, adding, “Following all the precautions has exhausted us,” she said. “The car was like this protective bubble where some people felt they didn’t need to follow the rules.”
The nature of distracted driving is changing. Over the last decade, the number of drivers talking on their phones has decreased, but the number of drivers “manipulating” their phones (texting, emailing, scrolling through social media) has more than doubled, according to Reagan. The associated risk is much higher than just talking on the phone. As in-vehicle screens and automated features demand more of drivers’ attention, visual-manual distractions pose the greatest threat. “Where the driver is looking tends to be related to where the driver’s paying attention, and the longer the driver looks away from the road, the higher their likelihood of getting into a crash,” noted Mueller. Consistent with IIHS research, the 2022 Travelers Risk Index on distracted driving, released March 30, found that more than half of people surveyed admitted to reading text messages or emails while driving.
No, your car can’t drive itself. “There are no vehicles available today that are fully equipped with self-driving technology, but some are designed in ways that can give the impression that they can drive themselves,” said Mueller. “These technologies are not meant to replace the driver whatsoever, and they can’t. Sometimes they will fail. Sometimes they will do things that the driver does not expect. The driver has to be engaged in order to prevent a dangerous situation,” she emphasized.
Automated driving assistance may increase driver distraction. A joint study between the Massachusetts Institute of Technology (MIT) and IIHS found that once drivers build up comfort and trust in their vehicle’s automated driving features, they become more disengaged. “The levels of disengagement went through the roof. It was night and day,” Reagan said. “We found much more distraction with the automation,” he added, explaining how having both hands off the wheel allowed drivers to fidget more with electronics as they developed trust in the driver assistance features in their vehicles.
IIHS’ newly established rating system aims to ensure drivers use the technology appropriately and remain attentive while driving. “The program sets minimum expectations for how these technologies should be designed so drivers are kept engaged when using them,” said Mueller. IIHS recommendations for in-vehicle automated driving systems include the ability to:
- Monitor for signs of driver disengagement.
- Clearly and rapidly communicate to the driver when disengagement is detected to return their attention to driving.
- Deter repetitive, prolonged disengagement through emergency escalation countermeasures.
- Require drivers to initiate or confirm automated lane changes.
- Verify the driver is looking at the road before automatically resuming from stop.
- Promote cooperation between the driver and vehicle through shared steering control.
- Require use of crash avoidance features and seat belts when using the automation.
When it comes to safety, crash avoidance technology rules the road. “It’s the most effective countermeasure that we have for distracted driving at the moment,” said Reagan. While the jury is still out on automated driver assistance technology, collision warning with auto-braking translates to a 50% lower rear crash rate.
No vehicle passed all the IIHS safety tests for partially automated cars. “No one is getting a gold star in every category, but some automakers are doing some things really well,” Mueller told us. “Everything that we are asking for is possible. This is the whole purpose of why we developed a ratings program to help push for designs that encourage proper system use.
Presented by the Travelers Institute, the Insurance Institute for Highway Safety, Safe Roads Alliance, the MetroHartford Alliance, the American Property Casualty Insurance Association, the Master's in Financial Technology (FinTech) Program at the University of Connecticut School of Business, and the Risk and Uncertainty Management Center at the University of South Carolina’s Darla Moore School of Business.
Slide, Text, Wednesdays with Woodward (registered trademark) Webinar Series
Joan Woodward appears on a video call. Text, Travelers Institute (registered trademark). Travelers logo.
JOAN WOODWARD: Good afternoon, and thank you so much for joining us. I'm Joan Woodward, and I'm honored to lead the Travelers Institute, which is the public policy division and educational arm of Travelers. Welcome to Wednesdays with Woodward, a webinar series that we created during the pandemic to help us get us all through it, where we convene leading experts for conversations about some of today's biggest challenges, personal and professional. So before we get started, I'd like to share a disclaimer about today's program.
Slide, Text, About Travelers Institute (registered trademark) Webinars. The Wednesdays with Woodward educational webinar series is presented by the Travelers Institute, the public policy division of Travelers. This program is offered for informational and educational purposes only. You should consult with your financial, legal, insurance or other advisors about any practices suggested by this program. Please note that this session is being recorded and may be used as Travelers deems appropriate.
Slide, Text, Ready to Take Over? Driver Distraction in the Age of Automation. Logos, MetroHartford Alliance, UCONN School of Business – MS in Financial Technology, Travelers Institute, IIHS HLDI Insurance Institute for Highway Safety, Highway Loss Data Institute, Safe Roads Alliance, American Property Casualty Insurance Association, University of South Carolina Darla Moore School of Business
To lead us through this topic today, I am thrilled to welcome a special guest host, my colleague Jessica Kearney. Jessica is the Assistant Vice President here at the Travelers Institute for the last 10 years and is really the driving force behind all of our initiatives. She shapes our agenda; she shapes our programming and our broader relationships in the communities. So, in 2016, Jessica led the creation of the Travelers Distracted Driving Educational Campaign-- Every Second Matters (registered trademark). And currently leads the Institute's auto safety and innovation projects.
She's spoken at dozens of events and conferences across the United States and is sought out as a leader on this topic. So, please join me as she'll be taking the helm today, my friend, Jessica Kearney. Jessica, take it away.
Jessica appears on the video call.
JESSICA KEARNEY: Thank you, Joan, and good afternoon, everyone. I'm so pleased to be here and I'm ready to jump right in. As many of you know, April is Distracted Driving Awareness Month and this year it's really as important as ever to recommit to roadway safety. According to the National Highway Traffic Safety Administration, more people were killed in motor vehicle collisions in the first nine months of 2021 than in any year since 2006. And just out this morning, our own Travelers Risk Index found that more than half of people surveyed still admit to reading text messages or emails while driving.
So, we have a lot of work to do and a lot of ground to cover. So, thank you for taking the time out of your day today for this conversation. We really hope there'll be lots of valuable takeaways. Also, a huge thanks to our co-hosting organizations you'll see here on screen. They are part of the process of helping us advance these really important safety conversations, and I'd like to recognize them.
The Master's in Fintech Program at UConn School of Business, the Risk and Uncertainty Management Center at the University of South Carolina's Darla Moore School of Business, the American Property Casualty Insurance Association, Safe Roads Alliance, MetroHartford Alliance, and the Insurance Institute for Highway Safety or IIHS. Many of you have probably heard of IIHS. Many of you have probably looked up their top safety picks, maybe the last time you bought a car or a vehicle. This is a fantastic, independent research organization that crashes vehicles (and lots of them) in their facility in Virginia, and it's all in the name of science and on the name of safety.
Slide, A picture of hands on a steering wheel. Text, IIHS creates safeguard rating for partial automation
Their publicly available crash test ratings have been moving the needle on auto safety and saving lives since 1959. A few weeks ago, you might have caught some headlines about IIHS. They have just rolled out a first-of-its-kind ratings program for partially automated vehicles, or if you think about it, vehicles that use things like cameras or radar or sensors to see the road and provide driver assistance features. And I put the emphasis on driver assistance and that's an important part that we're going to get into today. A critical piece of this new ratings program is how well each vehicle ensures that the human driver is actually still paying attention.
So, eyes up, hands on the wheel-- yes, that's still critically important. Today, for Distracted Driving Awareness Month, two of their leading researchers will take us behind the headlines. Interestingly, not a single vehicle available for purchase on the market today passes all of their new safety tests, so we're also going to dig into that. Now, before we start, we want to level set on why this all really matters. And I'm very, very pleased to introduce Emily Stein to help us put that in perspective.
Slide, Speakers, a list of pictures, names and titles
Emily is president of Safe Roads Alliance and entered the road safety field after her father, Howard, was tragically killed by a distracted driver in 2011.
Emily is a registered nurse, she holds a master's in public health, and she's a tireless advocate for roadway safety, both in her home state of Massachusetts and across the country. Our teams at Travelers had the privilege of working with Emily to honor her father through a video series we call “Unfinished Stories,” and I'm very pleased to welcome Emily for some opening remarks and then to share that video.
Slide, a still picture from the animated video of a man holding a little girl’s hand. Text, Howard's Unfinished Story, The Tree House
Emily, over to you.
Emily appears on the video call.
EMILY STEIN: Thank you so much, Jessica and Joan. And thanks to the Travelers Institute for all your efforts in helping to reduce distracted driving. I have worked in distracted driving awareness and prevention for nearly 10 years. My dad, who was the most important person in the world to me, was killed on his way home from work one evening in April nearly 11 years ago. The driver who hit him was programming her GPS while driving and didn't even see him. To help process my grief, I felt the strong need to talk about what had happened to my family in hopes of educating drivers about the power they hold when they get behind the wheel. The power to take away a life in an instant.
Now, as everybody did, we had great plans right before COVID hit in terms of driving safety education and outreach to the community. And I'm excited to say that, just on Monday, I was able to sit in a classroom full of third- and fourth-graders as they were being taught a program called, “Kids Speaking Up for Road Safety.” So, Safe Roads Alliance partnered with another organization called NDD.org to create this program with the goal of helping to give kids in the backseat the right tools to speak up if their adult driver was driving distracted.
So, this is one of the many ways that we're hoping to reduce this epidemic on our road, to help future drivers understand that distracted driving must be socially unacceptable, while also helping them to intervene and speak up if they felt unsafe in the car. I imagine many of you in this webinar have noticed the state of drivers these days since COVID. It seems that people are driving faster and more aggressively, they're talking and texting on their phones far more than before COVID, and then failing to stop for pedestrians who are waiting in the crosswalk.
COVID has clearly been tough on all of us in so many ways and following all the COVID restrictions and precautions has exhausted us too. And I think for some people the car was like this protective bubble where some people felt they didn't need to follow the rules or think about all the stressors that they had in their life. And unfortunately, as Jessica mentioned, this behavior has led to an unacceptable level of death and injury on the road the past two years. Just to shift back to the focus of this webinar, when thinking about car design, it's so important to think about how this new technology is both saving lives, but also causing some big concern.
We have these big infotainment screens in our car now that include backup cameras, which is a huge safety feature. But the screen also, in many cars, it doesn't lock out when the car is in motion and that's a big problem. Nobody should be able to order Starbucks from the driver's seat while driving 60 miles an hour on the highway. And then on the flip side, if this crash avoidance and driver assist technology was in the car that hit my dad 11 years ago, the lane departure warning would have sounded. The driver inattention would have been detected; an automatic braking might have been that last line of defense.
And my dad and hundreds and thousands of others might still be alive or not suffering from a serious injury today. So, I'm really hopeful that if we do this correctly with the federal regulations put in place now, these driver assists, and some partially automated features can indeed save lives instead of causing more distractions or driver disengagement. So, now we'd like to show a short version of the beautiful PSA that Travelers created about my dad as part of the series they made called, “Unfinished Stories.”
The video plays. Emily sits in a room.
My dad, Howard, was a self-taught carpenter.
An animation shows a man (Howard) in a hard hat working with wood in a workshop.
When he walked into the studio, he became an artist.
Howard hammers a nail and draws a line on a piece of wood with a pencil. He saws the wood.
He really had a vision.
Outside, Howard looks up. A treehouse appears superimposed on a tree. Emily sits with a young girl, who draws in a notebook.
I was pregnant with Evy, the first grandchild that my dad never got to meet.
In the workshop, Evy and Howard mark up a piece of wood together.
I imagine he would bring her into his workshop. I think he would have loved to do that.
Evy puts on a hard hat and then she and Howard sand a piece of wood together. Evy hands Howard a hammer.
Outside, they hold hands as they look up at the finished treehouse. Inside the treehouse, Evy runs and hugs Howard.
In the middle of the night, we got a phone call.
On the side of a road, Howard climbs into the driver's seat of a truck. In another car, a hand reaches for a handheld device sitting on the passenger seat. She looks up and swerves away from Howard.
My dad was driving home from work and a distracted driver swerved out of her lane, and everything in my body just said this was so preventable.
A framed picture of the real-life Howard, smiling and holding a piece of wood, sits on a table in a workshop. Text, Howard Stein, 1949 to 2011
Text, Distracted driving takes the lives of 9 people each day. Leaving their stories unfinished.
A red umbrella falls. Text, Let's end distracted driving. Every Second Matters (registered trademark). travelers dot com slash distracted driving
The video call resumes.
JESSICA KEARNEY: Emily, thank you for those opening comments and for all that you do for roadway safety and for that introduction to this very important topic. And with that backdrop, we're going to move right into our speaker presentations.
The Speakers slide reappears
As I noted earlier, we are thrilled to welcome two research scientists from IIHS. I'll introduce them now. They're going to share some opening presentations and then we'll reconvene for discussion with your questions, and on that, please feel free to drop your questions in the Q&A feature at the bottom of the screen at any point throughout the presentation and we'll get to them.
First up, we will have Dr. Ian Reagan. Dr. Reagan is a Senior Research Scientist at IIHS. Since joining in 2012, he's conducted research on crash avoidance technologies and driver distraction among other topics. Previously, he worked at the National Highway Traffic Safety Administration as a research psychologist. Dr. Reagan received a bachelor's degree and a doctorate in psychology from Old Dominion University. After Ian, we're going to welcome Dr. Alexandra Mueller. Dr. Mueller joined IIHS in 2017 and conducts research on the usability and safety of driver assistance technologies and driving automation.
She received her bachelor's and master's degrees in psychology from the University of Guelph, and a doctorate in psychology from the University of Western Ontario. She completed postdoctoral fellowships at McMaster University and Cincinnati Children's Hospital Medical Center. And with that, let's kick it off with Dr. Ian Reagan to begin the presentations. Ian, over to you.
Ian appears on the video call. The title slide reappears.
IAN REAGAN: Thank you, Jessica. Very honored to be here on this webinar and I'd also like to thank Emily for her work on this. A lot of times in the work that I do, research, I get lost in the numbers and hearing a victim who's been affected by a distracted driving crash really sort of brings things home. So, I'll go ahead and go into my presentation. If you want to start with the first slide.
Slide, Text, Percentage of drivers stopped at controlled intersections observed using cellphones during a given daylight moment, 2011. A graph shows two lines, with holding phone to ear gradually decreasing, and manipulating phone gradually increasing.
I guess I'll start by circling back to the points that Jessica and Emily highlighted about this increase in traffic fatalities during these years of the pandemic.
And one of the reasons that it's so upsetting is that we know that driving, overall, was down yet fatalities were up. And at the same time, we believe that distraction is part of the equation here, but we also know that the numbers on distracted driving fatalities really underestimate the problem in a big way. Just a quick example, in 2020 only 8% of all fatalities were attributed to distracted driving, and that just seems low. And I think that the underestimate reflects the difficulty, the challenges of identifying whether or not distraction was present in a crash.
If you consider, for example, single vehicle crashes where the driver was the only person involved and they were fatally injured, how is an officer who's completing the report to determine that distraction was a factor? These numbers that you're looking at are a lot more reliable, and these are based on observations that are conducted on the roadside. And they represent the percentage of drivers in the U.S. who are observed using their phones while stopped at an intersection. And over the last year, or the last decade, what we're seeing are year-by-year trends that sort of highlight really big change and fundamental change in how drivers are using their cellphones.
So, we're seeing fewer drivers talking on their phone over the last decade, but during that same time period, we've seen the percentage of drivers who are manipulating their phones has more than doubled. And the reason that is concerning to us is we know that cellphone manipulation has a much higher crash risk associated with it than just talking on the phone, and that goes back to a large body of research that has indicated that it's the visual manual distractions that are the big factor in distracted driving crashes. And so, I understand that there are some Canadians on the webinar today.
And I contacted Transport Canada and they indicated that they're seeing the same numbers. So, this is a really concerning trend and so I wanted to move on to talk about some of the countermeasures that are available for addressing distracted driving, as we know it to be right now.
Slide, A timeline from 1997 to 2017. Text, Timeline of distracted driving countermeasures
And one of the points that I like to emphasize when talking about distracted driving is that it isn't just
cellphone use. However, at the same time there's no denying that cellphone distractions get the bulk of the attention. So, I’m going to spend a few minutes talking about the different countermeasures here on the slide and see that the first point, in fact, is about cellphone use.
There was a study in 1997 that came out just as cellphones were starting to proliferate, and they indicated having a cellphone conversation was associated with a four-fold increase in crash risk. And so, it wasn't long after that, that New York became the first state in the U.S. to ban hand-held cellphone conversations for all drivers. And since then, fast forward to 2022, we have 24 states in the U.S. that have some ban on hand-held phone use. Many focusing on hand-held conversations. And there are 48 states plus the District of Columbia that have laws that are specific to banning text messaging.
So, the thing is that we have all these laws, but it's unclear as to whether or not they have had an effect in terms of reducing crashes. There have been a number of evaluations that have looked at this. Some have concluded that crashes went down after the laws were put in place, but there have been other studies that have found no effect, and a couple that have actually reported that crashes went up in the jurisdictions after laws were implemented. And so, one of the last things I'd like to say about the legislation and enforcement of those laws is that these states have enacted the laws as technology has advanced rapidly over the past two decades.
And you can see on the timeline, first iPhone came into play in 2007. Soon after that, you started to see software apps that were introduced, so that the software was optimized for use on smartphones, and Snapchat is a perfect example of a social media platform that's really designed for use on the smartphone. And that gives context to the increase in cellphone manipulation that we've seen over the past decade. And so, the states that have enacted laws more recently, they're starting to use more comprehensive wording.
So, they're banning-- the laws are banning drivers from simply holding a cellphone while the vehicle is in motion. So, it'll be interesting to see in the next couple of years after these stronger laws have been in place to see whether or not crashes have been reduced as a result. So, in addition to the technology advances that have gone on with cellular technology, Emily kind of pointed to the way things have changed inside the vehicle. And that brings me to the next countermeasure that I wanted to talk about, and that's designing vehicle interfaces, so they limit visual-manual distraction.
And the logic behind NHTSA's effort to release guidelines for reducing visual-manual distraction with vehicle systems-- the logic was, drivers are at times going to need or want to access their in-vehicle systems. And they should be allowed to use these systems to complete tasks, provided those tasks have acceptably low levels of visual distraction. So, there was a lot of effort that went into defining what those allowable levels of visual-manual demand should be, but the problem, the big limitation with the NHTSA guidelines is that they're voluntary.
So, there's no sort of penalty to automakers if they implement an infotainment system that lets drivers conduct tasks that introduce more visual demand than would be recommended by the guidelines. So that's a little bit disappointing. That brings me to the last two countermeasures on the slide that I want to talk about, crash avoidance technology and cellphone blockers. So just kind of hit on cellphone blockers real quickly. These applications, they do have some potential to them. But the way they've been implemented now, they're limited in the fact that they only become active if drivers opt in to use them.
So, there are a number of blockers out there. Apple introduced one in 2017 called, “Do Not Disturb While Driving,” and it was of interest to us because this was the first one that was implemented on a wide scale where all iPhone owners were told that the app was installed, but they were also asked if they wanted to try it out. And so, the app was turned on, but only if they agreed that they wanted to try it. And so, we did a study and found that only 20% of drivers who owned iPhones had the app set to block notifications automatically whenever it determined that they were driving a vehicle.
And so, we see the potential in these cellphone blockers is for the cellphone producers to use-- instead of an opt-in strategy where they ask the driver if they want to try it, to use an opt-out strategy where the owner is told that the app is there, but it's turned on to activate automatically. But they can go-- the owner can go in and turn the blocker off if they so choose. So, the last countermeasure I wanted to talk about was crash avoidance technology. And I wanted to hit on it last because I believe it's the most effective countermeasure that we have for distracted driving at the moment.
And a couple of reasons why I say that-- one being, what we know about distracted driving is that the lion's share of distracted driving crashes are due to visual distraction. We've done a lot of work, Insurance Institute, to show that vehicles, for example, with forward-collision warning with auto brake, have a 50% lower rear-crash rate than vehicles with no front-crash prevention system. These crashes that tend to happen because drivers look down and then something happens when they look away. These systems that we've evaluated are reducing those crashes.
And the other point that I like to make when emphasizing the effectiveness of crash avoidance for distracted driving is the fact that these crash avoidance systems are blind to the source of distraction, whereas the other three countermeasures that I talked about focus specifically on cellphones or in-vehicle displays.
Slide, Text, Partial driving automation presents a new set of concerns for driver attention
So, moving on to where we are heading in terms of the other vehicle technologies that are becoming available to us. We see the development of partial driving automation as really an inflection point with highway safety in terms of it having a lot of potential to improve safety, but at the same time there are a lot of concerns about what's going to happen with driver attention.
Slide, pictures of assistive technologies labeled levels 0 through 5. Text, Vehicle driver assistance and automation is a spectrum
And so, if you want to move on, in a little bit of time I have left, I want to summarize where we are in terms of what's available to consumers with vehicle automation right now, and then highlight our concerns about drivers using the automation. And so, some years ago, the SAE, the Society of Automotive Engineers, developed a classification system for automation, which you can see you have crash avoidance systems at level 0 where the automation is only active and when it, in momentary instances when it detects that crash is imminent, all the way up to level 5 automation where the driver can basically sit in the car and go to sleep and wake up when they arrive at their destination.
But our concerns, and what Alex is going to be talking about in a minute, are levels 1 and 2 driver assistance systems. So, if you were to go into a dealership today, you could buy vehicles with level 0, 1, or 2. There's currently no consumer vehicle available with levels 3 through 5 technology. So, with driver assistance systems, where our concern is that the systems provide continuous support for some aspect of the driving task, but there's still a very large role for the driver when they're using these level 1 and level 2 systems.
And our concern is more with level 2 than level 1. So just briefly, level 1 systems you're talking about adaptive cruise control by itself. ACC takes care of vehicle speed and following distance. The other level 1 technology is lane centering, and that's just continuous steering to control the lateral positioning of the vehicle. And when you combine the two, you have a level 2 partial driving automation system. And our concern is that when you're using level 2, the driver is fully responsible for the object detection and response task.
So, for example, you're driving down the interstate. There's a wooden pallet in the middle of the lane, the level 2 system is not going to detect and avoid that. That's completely on the driver. The systems will also fail-- can fail at a moment's notice and the driver has to be ready for that. And so, at this point during my presentation, I was hoping to play a media clip that really did a great job of underscoring our concern that drivers are just going to check out with this technology. It was news footage that literally showed drivers asleep at the wheel when they were using this level 2 automation.
You can go on YouTube and see plenty of videos of drivers misusing this technology.
Slide, two pictures of drivers behind the wheel. Text, Driver disengagement study: hands-off-wheel and distraction during a 4-week field trial.
But moving on to our next slide, we've done a number of studies on our own, or in collaboration with other researchers, that has basically shown our concern that the driver's level of distraction is going to be greater when using this partial driving automation than it is when they're driving manually. So, in this study that we did in collaboration with MIT, we followed a small group of drivers around for a four-week period.
Everybody drove a vehicle with partial driving automation, and we looked at how disengaged they were while driving. So how often they had their hands off the wheel or were engaged with using a phone or the in-vehicle systems. And we compared that disengagement levels when they were driving manually versus when they were using the automation. And during the first two weeks of the study, we didn't see any differences between disengagement levels when they were driving manually versus using the automation--
Slide, a picture of a driver engaged in a video call on her smartphone. Text, Drivers let their focus slip as they get used to partial automation, November 19, 2020
but moving on. When we got to that second two weeks, the drivers had built up some comfort using the partial automation and began to trust it.
And what we saw is that their levels of disengagement went through the roof compared to when they were using the automation during the first two weeks. And so, when we compared the distraction levels when they were using automation in weeks three and four to manual driving and weeks three and four, it was night and day. Much more distraction with the automation. And so that brings us to the last study that I wanted to tell you about.
Slide, a bar graph. Text, Which behaviors would you consider safe while using? Percent of drivers in 2018 IIHS survey
Not only do we see this change in behavior, but things as basic as what the automation is named can really lead drivers to have misperceptions as to what is OK to do when you're using the technology.
So, in this study, what we did was ask drivers if they thought it would be safe to do all these different behaviors that you see across the bottom of the figure. And we asked them if it would be safe to do so when they were using a recent automation system that helped the drivers keep the vehicle in the lane, control their speed and following distance. But what we did is we told them that the technology was one of the five brand names that you see across the top of the figure. And it's important to note that out of those five systems, only the Super Cruise system was designed to allow drivers to take both their hands off the wheel.
Yet, when the drivers heard autopilot, they were much more likely to say that it was safe to take your hands off the wheel, talk on a cellphone, text, watch a movie or even take a nap. So, these findings really just led us to conclude that something needs to be done about ensuring that if drivers are going to use these level 2 systems, that they're attentive when doing so because of this tendency to check out. And that's what led us to develop the safeguard ratings that Alex is going to tell you about now.
Alexandra appears on the video call. Slide, text, IIHS ratings program on safeguards for partial driving automation. Travelers webinar. Alexandra Mueller, Research Scientist.
ALEXANDRA MUELLER: Thanks very much, Ian. So, before I begin, there are some things I want to talk about. First and foremost, as Ian said, there are no vehicles available to buy today that are equipped with self-driving technology. But some of these systems are designed in ways that can give the impression that they can drive themselves. Now, the problem with this, as Ian just showed you, is that sometimes people do things other than drive when they use these systems. So, we created this ratings program to require safeguards in these systems to prevent people from treating the technology as self-driving.
Unlike crash avoidance features, partial driving automation does not have a clear empirically supported safety case or benefits, and the program is not an endorsement of the technology of partial driving automation. So, a good way to think about this program is that it sets minimum expectations for how these technologies should be designed so that drivers are kept engaged in driving when using them. And I'll go into each category and the reasons behind the safeguard requirements now.
Slide, a bullet point list. Text, Driver disengagement. Indicators that the driver is out of the loop.
So, first of all, these systems have to monitor what the driver is doing to ensure that the driver is always ready and able to intervene at a moment's notice. Functional testing has shown that these systems frequently encounter conditions that they cannot handle that require the driver to very rapidly intervene. Sometimes these conditions are so mundane to the driver that the driver actually might not expect this problem and the struggle to even happen. And we know that where the driver is looking tends to be related to where the driver is paying attention to.
And the longer the driver looks away from the road, the higher the likelihood of getting into a crash. But what the driver's hands are doing is also important because hands-off wheel behavior is related to looking away from the road as well as distraction. If the driver's hands are occupied, that will impair one's ability to steer. So, next slide, please.
Slide, illustrations of a driver's eyeline while behind the wheel. Text, IIHS category 1: Driver monitoring. Must monitor driver for signs of disengagement
So, our first category of the ratings program is to ensure that the system monitors where the driver is looking and what the driver's hands are doing to ensure that he or she is fulfilling their role when using the technology. Next slide, please.
Slide, a bullet point list. Text, Communicating to the driver. How can the driver be brought back into the loop?
Now, but what happens if the driver starts to do things that he or she is not supposed to, such as texting? The system has to rapidly begin communicating to the driver to get them back into the loop of driving. This type of communication is called an attention reminder. Clear, escalating, multimodal communication improves the likelihood of the driver detecting and responding to the alerts. And timing is also of the essence. Non-visual forms of communication help to capture the driver's attention, and the visual information that's provided helps to tell the driver why the alerts are happening and what the driver has to do.
The instrument panel is one of the first places people look when they get a non-visual alert. Moving on.
Slide, a bullet point list. Text, IIHS category 2: Attention reminders. Rapid timing and utilization of multiple channels of communication.
Thanks. So, our second category of the program requires that the system uses attention reminders that escalate rapidly in urgency. And that escalation should incrementally add modalities or independent channels of communication at each stage, so that it escalates to a minimum of three unique methods of communication simultaneously, thereby helping to maximize the chances of the driver detecting and responding to the alerts. But what happens if the driver doesn't respond? Moving on, please, to the next slide.
Slide, a bullet point list. Text, What if the driver does not respond? Countermeasures
Well, the worst-case scenario would be for the system to continue to drive itself or try to drive itself anyway, or to simply switch off. Remember, these systems cannot drive themselves, but if the driver isn't in the loop and if the system switches off, that means the vehicle is traveling at speed completely uncontrolled. A better scenario or a lesser of these evils, so to speak, is to have the system do a controlled slowdown. Emergency escalation countermeasures, such as restricting driver access through a lockout mechanism, helps to deter prolonged and repetitive system misuse.
Slide, a bullet point list. Text, IIHS category 3: Emergency escalation
So, our third category-- moving on, thank you-- requires the use of these countermeasures if the driver does not comply to the attention reminders. The vehicle should perform a controlled slowdown or a stopping procedure. And it should also initiate an SOS call because at that point, it is very possible that the driver is incapacitated and needs help. If the driver hasn't responded to the attention reminders and if the driver is enduring a vehicle slowdown on highway conditions. But once the vehicle slowdown procedure has begun, though-- and of course, the driver can resume control at any point during this process.
Once the slowdown procedure has begun, though, it is clear that the driver has been misusing the technology beyond what it is designed to do. So once the driver either resumes control or the vehicle has been brought to a stop or a crawl, the driver should no longer be able to access the automation for the remainder of the drive. Now, the next two categories have to do with responsible application of additional automated functionality. Next slide, please.
Slide, a bullet point list. Text, Responsible application of automated functionality,
An example of this functionality is automated lane changing, where the vehicle will make the lane change maneuver on its own without requiring the driver to steer.
The difficulty with these automated functionalities on top of the traditional level 2 support with ACC and lane centering combined, is that it starts to become difficult to convey to the driver that these systems are not self-driving when they can perform complex maneuvers without the driver being physically involved. Remember, all of these systems require the driver to supervise and ensure that the maneuvers are safe to perform. And, if it's not safe, the driver has to intervene very rapidly. But it becomes difficult for even an attentive driver to anticipate and respond to the vehicle's behavior if the vehicle is making these complex maneuvers on its own. Next slide, please.
Slide, a bullet point list. Text, IIHS category 4: Automated lane changing
So, the fourth category of this program concerns automated lane-changing functionality. If an automaker wants to offer this functionality in their systems, the feature must require the driver to either physically initiate the maneuver or to give some sort of physical confirmation before the lane change maneuver will be performed. That way the system has some sort of physical indication from the driver that they are still in the loop. Also, as per our requirements, it will only be acceptable to offer automated lane-changing functionality if the system performs well in all the other safeguard categories.
Because at that point, if the system has performed well in all the other categories, the system is demonstrating that it has safeguard designs to keep the driver engaged. Next slide, please.
Slide, a still picture from a video of a driving test. Text, System behavior without driver involvement. Driver distraction is prevalent while at stop and slow speeds
Now, some of these systems can automatically resume speed from stop. For example, in a traffic jam. But the problem with this is that driver distraction tends to be prevalent at slow speeds or at a stop, and this video is going to show you why the driver needs to be alert during the auto resume situations.
A video plays. A car slows to a stop while approaching another vehicle. A dummy stands on the side of the road.
So, you can see here that the pedestrian dummy is going in front of the vehicle in just a minute.
The dummy walks in front of the car and stops.
It's in front of the vehicle. The lead vehicle ahead begins to move, ACC auto resumes and unfortunately hits the dummy because the system did not recognize its presence. Next slide, please.
Slide, a bullet point list. Text, IIHS category 5: ACC auto resume
Again, this is a good example of what Ian was talking about in that sometimes these systems will do things that the driver does not expect. And sometimes they will fail. So, if these systems are to be equipped with this auto-resume functionality, our fifth category requires that they must verify that the driver is looking at the road before the auto resume will activate.
Also, if the vehicle has been at a physical stop for an extended period, the auto-resume functionality should time out and require the driver to give some physical indication to the system when he or she is ready for the system to begin moving the vehicle. This could be very simply done, for example, through a button press or through the throttle pedal press. Next slide, please.
Slide, Close-up of hands on a steering wheel alongside a bullet point list. Text, Keeping the driver involved. Proactive design strategies through shared control.,
Now, as these systems are driver support features, the driver should be an active participant in the driving task. And this can be facilitated through shared control with the lane-centering support. Specifically, through sharing steering control.
People are better at proactively controlling the vehicle when they have information feedback through the steering wheel. Shared control with the steering also helps to convey a driver's sense of autonomy in the interaction, so that it becomes a matter of driver working with the machine instead of driver versus the machine. Now, vehicle position cues through the steering wheel can also help a driver predict the automation steering behavior. This also allows the driver to detect problematic situations more quickly and intervene accordingly.
However, some systems out there switch off the automation support whenever the driver steers, and the problem with that sort of design is that it can be perceived as a form of punishment for trying to participate in the driving. Next slide, please.
Slide, Close up of hands on a steering wheel. Text, IIHS category 6: Cooperative steering assistance
So simply put, our sixth category requires that systems must not switch off whenever the driver steers. This is meant to promote cooperation between the driver and the vehicle and keep drivers engaged in driving. Next slide, please.
Slide, another image of hands on a steering wheel alongside a bullet point list. What features should be used with these systems? Established vehicle technologies
Now, as I mentioned before, there are no clear safety benefits established yet for partial driving automation, but there are clear safety benefits for crash avoidance features.
As Ian said, many of the types of crashes that automatic emergency braking and lane departure prevention address result from distraction, such as rear-end and lane-drift crashes. Also, seatbelts have a clear safety benefit for mitigating injury severity in the event of a crash. Next slide.
Slide, a bullet point list. Text, IIHS category 7: Safety features
Now, very simply, this last category of the program requires that in order to use the partial automation, these crash avoidance features must be on AEB and Lane Departure Prevention. Now, drivers also have to be buckled in order to switch on the automation, but if for whatever reason the driver becomes unbuckled after switching on the automation--
For example, let's say the driver wants to take off their jacket, this should trigger the attention reminders to begin. And if the driver does not re-buckle in time, it should begin the emergency escalation process. Because at that point, the driver is doing something unsafe. Thank you very much.
Jessica reappears on screen.
JESSICA KEARNEY: Alright. Alex, Ian, thank you both for those fantastic presentations. We're going to reconvene now and get into our moderated discussion, and importantly, our audience questions. That was fascinating and lots of great takeaways there. And Alex, I just wanted to continue on some of the thinking and some of what you just talked about. So, David Harkey, President of IIHS, gave a good quote recently that said, “Partial automation systems make long drives seem less of a burden, but there's really no evidence that they make driving safer.” And I think you just echoed that in your presentation.
So, is that the headline here and, if so, how much farther do we actually have to go on this automation journey before we really start to see some of those real safety benefits kick in? I know we talked about a crash avoidance and some of the others that are showing real benefits today, but where are we on this journey would you say?
ALEXANDRA MUELLER: Well, there's an important distinction to make right now about partial driving automation and crash avoidance features. Firstly, as you just acknowledge, crash avoidance features are available on many vehicles today-- far more than partial automation. And we know that crash avoidance features work. They help to mitigate and reduce crashes. And right now, as we said earlier, the evidence is mixed around partial driving automation. But it is possible with better safeguards implemented into partial driving automation that's available now, they might have some safety benefits in the future.
This is the whole purpose of why we have developed a ratings program around it. But fully self-driving vehicles will have their own challenges and they won't necessarily be a silver-bullet solution even when they become available to consumers. But a very important thing to note is that there are many things that we can do short of self-driving vehicles that could produce immediate safety benefits right now. For example, pedestrian safety measures, infrastructure changes, lowering speed limits. These changes can be put into effect right now. And with the benefits that we're seeing with crash-avoidance technologies, more seatbelt adherence, we can be saving a lot of lives.
JESSICA KEARNEY: That's great, that’s great. And I know you talked a little bit about the scale at which some of these functions are available today, and Ian, I want to pull you in here. What percentage of new cars on the road today are equipped with these driver assistance features? I know that there's several of them available. And, as we think about the whole fleet, if you think about the whole fleet across the U.S. or Canada, all the passenger vehicles available-- what do we know about how long it takes some of those safety features to actually trickle down so that it's on the majority of vehicles on the road?
IAN REAGAN: So, two-part question. That second question, I think, is a little more straightforward to answer. So, the short answer is it takes about 35 to 40 years for a technology to really become ubiquitous in the fleet. And there's some-- it varies depending on the technology, but just as an example, rear cameras. They became available in 2002. In 2011, I think it was, there was a mandate that all cars have a rear-view camera. Forward to 2020, we have 45% of the U.S. fleet of vehicles estimated on the road to have a rear-view camera.
And it's not until 2040 that our sister organization, HLDI, estimates that we'll be at 95%. So, it varies a little bit depending on the technology. As far as the level 2 driver assistance, HLDI's actually preparing to release some data on that. They've been developing a methodology to come up-- help them estimate that one of the tricky things with the driver assistance level 2 systems is that some automakers-- and Honda and Toyota are good examples of automakers that have vehicles with adaptive cruise control.
They also have lane centering, but they don't market the system as a level 2 system as Tesla has with Autopilot or Volvo has with Pilot Assist, and so forth. And so, how do you treat a scenario like that? But HLDI is working on that and they will have some estimates in the next few months and expect that initially it'll be a very low percentage of vehicles on the road with level 2 systems.
JESSICA KEARNEY: Yeah, so, we're definitely looking at some time here before this does have widespread rollout. And then, Alex, I want to pick up on something that I teased in the opening remarks, which is that there's not currently a single vehicle on the market according to IIHS that meets all of the criteria for your new rating program. Do some of the systems perform well, though, in some categories? Can you give us some context there?
ALEXANDRA MUELLER: So, that's actually a great point. Not everyone-- there is no gold star. No one's getting a gold star in every category. But some automakers are doing some things really well. And what's important is that every category-- the way we've developed this program is it's based on what we know. And it's also based on the technology that's available today.
So, the fact that we're seeing automakers do well on some categories means that everything that we're asking for is possible, it's achievable. Unfortunately, no one is doing everything right now and that is the whole purpose of why we've developed this program. To help push design solutions around the safety component, specifically around promoting proper system use.
JESSICA KEARNEY: Now, what can we expect-- I know, I believe IIHS hasn't issued its first set of ratings yet, but you're planning to. Can you give us a window of what we might expect, what we might be looking at, what that's going to look like when it actually does come out?
ALEXANDRA MUELLER: So, we're hoping to release a first round of ratings by the end of this year. Unfortunately, due to chip shortages, which I'm sure everyone has heard about and supply chain issues, even getting these vehicles equipped with-- typically, it's the highest trim level. Typically, those vehicles come with all the bells and whistles. So, right now, getting access to all the relevant vehicles that we want to is what's been delaying us. However, our first round of ratings, what we plan to release, will contain a sample that is representative of the diversity of the systems that are currently available in the market.
And we're hoping that-- what we've seen with other ratings programs is that automakers pay close attention to them. And we've seen automakers rapidly improve. For example, with crash test improvements, crash avoidance performance, and we hope and expect to see the same with this program.
JESSICA KEARNEY: That's great. And if you're a consumer, so if you're looking at the first set of ratings that IIHS issues and you're trying to decide on making a purchase, what would you recommend, or what would you recommend that consumers focus on or consider if they're in the market for one of these vehicles?
ALEXANDRA MUELLER: Well, we hope that this ratings program will provide consumers with guidance for the safest implementations of these systems, but until our ratings come out, consumers should really think about what they really need. They should not think that these systems can allow them to do other things when behind the wheel, such as do work on their daily commute. None of these systems are self-driving and consumers need to understand that these technologies are not meant to replace the driver whatsoever. And they can't.
They are merely driver support or assistance features. And as I'm sure many of us who have driven these technologies know that sometimes they will fail, or even if they don't fail, sometimes they will do things that the driver does not expect. And when those situations happen, it can be very abrupt, and the driver has to be engaged in the driving task in order to respond in order to prevent a dangerous situation from happening.
JESSICA KEARNEY: And Ian, I want to pull you in here and Alex just alluded to this. IIHS is known for helping move the needle on safety through your ratings programs in the past. Can you give us any examples in the past where that's been particularly successful? Just kind of helping us put a framework around what we might potentially expect to help move the needle this time around?
IAN REAGAN: Sure. And you're absolutely right. Alex did allude to that with our crash avoidance and crashworthiness programs. There is evidence there. So just as an example, our first crash test was moderate overlap front crash test. Crash a car into a barrier. I think we started doing that test in 1995, and at the time I think there was less than 20%-- the first five or six years that we did the test, less than 20% of the vehicles we tested got good ratings. A few years later, we were up at about 50%, 60% receiving good ratings.
And, after 2012, we haven't tested or put a vehicle through with the moderate overlap test that's received anything other than a good rating. So, the automakers make these changes, but then the other key piece that is I think even more important is we've gone out and looked at real-world data. And we're able to show that the drivers who were involved in frontal crashes, they were driving a vehicle with a good rating on this test. They were 46% less likely to be involved in a fatal crash compared to somebody who was driving a car that was rated poor.
And we have in just about all our tests that we do, you can see the progression going from majority marginal or poor ratings to majority good or acceptable ratings with time.
JESSICA KEARNEY: And I think that's a good place for just a quick plug about just kind of the ratings that IIHS already puts out. If you're joining us today and you've not gone to the IIHS website and looked up your own vehicle, you can visit their website, put in your make, model, year of your own vehicle. And it will show you the crash-test rating and it's just a really great educational resource. We'll drop that in the chat, but it's a great resource for everyone. And then, Alex, I want to ask you too. What was most surprising when you're working on these ratings? Was there anything that came out of it that you weren't expecting?
ALEXANDRA MUELLER: So, there are a few things that really pop out to me. And probably, the first one is just how genuinely interested people from all over the world are in this program and in the topic of partial driving automation. This is a very nascent field. We're only now just starting to begin to understand how humans interact with these technologies and how that interaction evolves over time and changes as the technology itself changes.
The fact that so many stakeholders want to talk about the technology and the innovation around it represents a unique opportunity to do research and find solutions to make these technologies as safe as possible to use. But another thing that has kind of popped up over and over again is this strange misconception that sometimes is touted that the push for safety comes at the cost of innovation. And that's just simply not true at all. By requiring safety to be a paramount requirement around which innovation is created, it raises the bar for everyone.
And raising the bar is not a bad thing, although we fully recognize it makes it more difficult to design around and to design for convenience alone. But the thing is it's not acceptable to say, oh, it's too difficult and that'll do, when it comes to convenience-based designs that undermine safety. By requiring designs that incorporate safety, a safety focus, the innovation will emphasize empirically driven design solutions that are often ingenious, and in fact, give the manufacturers a competitive edge. And on the other side of the coin, it's bad for business to implement unsafe technology that puts the customer base at risk. So, there's also a long-term financial incentive to create good design.
JESSICA KEARNEY: Yeah, that's great. So, it's not one or the other. The two can and should work together to your point. And our last few minutes in here, because we're almost at the top of the hour. I do want to get to some audience questions. And Ian, I'll ask you we've gotten actually a lot of questions from people asking what's the most compelling piece of evidence or data that I can give someone-- a loved one, a colleague-- about distracted driving to try and change their behavior? Is there anything that pops out to you or any resources that you'd like to share with the audience today?
IAN REAGAN: I would say probably the most compelling data about distracted driving comes from a method of studying driver behavior called naturalistic driving research. And it basically put cameras and all other sorts of sensing equipment inside a vehicle, and you get as big a sample of drivers as you can afford to get, and you just follow them driving in their normal everyday life for as long as you possibly can. And the idea is that over time, with enough drivers, you're going to see a lot of near crashes and safety-critical incidents that you can go back and review the footage and determine what was going on.
And there have been scores of studies that have been based on that research. And time and again, they emphasize the role of visual-manual distraction. It's just really representing the overwhelming share of these crashes. And I'm not trying to say that cognitive distraction is not a risk, but if we could just capture visual-manual distractions, then we would really make a huge change on distracted driving.
JESSICA KEARNEY: Yeah. Thank you for that. We also got several questions just around speeding, right? And some of the changes we have talked about the changes in roadway behaviors during COVID. I don't know if either one of you might like to address kind of the importance of speed and as a factor in these collisions, and just as a factor in roadway safety where we are today.
IAN REAGAN: Yeah, I can give you my thoughts on that question. It is a huge issue. I think it's overlooked. Major factor-- and there's this new approach, sort of a philosophical approach to traffic safety. It's called the safe systems approach. And one of the key components in the safe systems approach is safe vehicle speeds. It's an overarching thing. The safe systems approach says, let's take this-- it's a multifaceted approach where we have to focus on making drivers as safe as possible, making vehicles as safe as possible, designing roadways so they focus on safety over efficiency and throughput.
But you have those three components, but another separate component is safe speeds. It's almost-- the way that I think speed is underestimated as a role in crashes almost to the extent that distracted driving is. And something needs to be done about it.
JESSICA KEARNEY: Great. Thank you for that. We are at the top of our hour. I want to send my enormous thanks to Emily for her opening comments, to both of you, Dr. Reagan and Dr. Mueller for your time and your expertise today. We hope everyone joining, we hope you had some good takeaways from today's program and we also hope that you'll pay it forward for Distracted Driving Awareness Month and share some of what you've learned and some of the safety tips for those around you. Watch out for the replay. This session will be sent to you in your inboxes very soon.
Slide, Text, Upcoming Webinars, Register, travelers institute dot org
And also, I'd like to invite you to join us on some of our other upcoming programs, which you can see here on screen. So, we're going to continue the conversation. April 6, 60 Minutes in Personal Insurance. We have two of our personal insurance leaders, Michael Klein and Loree Toedman. We'll follow that up on April 13 with a session on LinkedIn and how you can grow your LinkedIn (LinkedIn is a registered trademark of LinkedIn Corporation and its affiliates in the United States and/or other countries.) presence and really enhance your brand online.
And then we'll follow it up on April 20 with The Future of Fighting Insurance Crime and we're going to feature Travelers’ EVP and Chief Claim Officer, Nick Seminara, as well as the President and CEO of the National Insurance Crime Bureau, David Glawe.
Slide, Text, Watch replays, travelers institute dot org. Connect, LinkedIn, Joan Kois Woodward. Take Our Survey, Link in chat. Hashtag, Wednesdays with Woodward
So, you can register for any of these programs today by visiting us at travelersinstitute.org. If you haven't already, please take a moment and fill out our survey about today's program. Thank you again for joining us today for this important roadway safety conversation and have a great afternoon. Thank you.
Text, Travelers Institute. Travelers logo. Travelers institute dot org
Join Our Email List
Get on the list to receive program invitations, replays and more.SIGN UP NOW