Where is the female crash test dummy? | Astrid Linder and Emily Thomas

ET: 00:00

Women are not just small men. It's not as though we are exactly the same and you've just kind of like created like a fun-sized version and that's good enough. You know, there are significant differences that impact our biomechanical response in a crash.

TS: 00:20

Welcome to Made For Us, the show where we explore how intentional design can help build a world that works better for everyone. I'm your host, Tosin Sulaiman.

In the book Invisible Woman, Caroline Criado Perez explores how a gender data gap often leads to products that aren't designed with women in mind, like virtual reality headsets, smart watches, even piano keyboards. ‘Designers may believe they're making products for everyone,’ she writes, ‘but in reality, they're mainly making them for men.’

Another area where a gender data gap still exists is in car safety design. Crash test dummies used to assess the safety of new cars are modelled on the average male, even though both men and women drive and male and female bodies react differently in a crash. Studies have also shown that women have a higher risk of injuries like whiplash in certain types of crashes.

To understand why this gap exists and what's being done to fix it, I reached out to Emily Thomas, Associate Director of Automotive Safety at Consumer Reports in the US, who you heard at the start of this episode, and Astrid Linder, Professor of Traffic Safety at the Swedish National Road and Transport Research Institute. She led a team that developed the first crash test dummy representing the body of an average female. But three years after that dummy was released, here's where things stand.

AL: 01:36

The average female is not represented, she's not required, and therefore she doesn't exist as a commercial crash test dummy. The regulation tells you that you have to use the model of an average male, full stop. What needs to be done is after the full stop, it should say, and average female.

TS: 01:56

In this episode, we'll look at how close we are to achieving so-called safety equity and why this will become even more critical with the rise of autonomous vehicles. Here's Emily Thomas kicking off the conversation. I asked her how car safety has evolved since she's been working in the field.

ET: 02:13

I've been in the automotive safety field now for at least 15 years and we have continued to make progress. We're continuing to see traffic fatalities and injuries go down. We're not there yet, right? The goal if you talk to anybody who works in this field, like we all want to see zero. We want to see zero deaths, zero injuries. And so there's still work to be done. But even in the time that I've been in the field, seeing how we have made advancements in front passenger safety and now focusing on the back seats, recognizing that we've created this gap in safety within the vehicle and trying to make sure that we're taking proven technologies and now moving them into the back seats so that we're having better equity in the vehicle regardless of where you sit. So there's been a lot of progress and there will continue to be progress because there's tons of great research and people who are really dedicated to this cause. So that's been really exciting.

TS: 03:10

So, I mean, it is good news because I guess cars are safer than they've ever been for everyone.

ET: 03:18

Absolutely. Just our understanding of how to better protect the occupants within the vehicle has evolved over the years. so cars are so much safer than they ever were before. And we've implemented so many new technologies within the vehicle. We're continuing to optimize them for various demographics. And even the fact that cars can now try to prevent crashes. If you think about technologies like automatic emergency braking, if the vehicle is now able to sense a car in front of you that has suddenly come to a stop or slowing down and it's able to adjust your vehicle speed and apply the brakes to prevent you from having a rear-end crash. That's amazing. That's come a long way. So there's just more and more that is being applied to the vehicles and we're just making them better so that we are able to reduce the number of road traffic deaths and injuries.

TS: 04:10

I wanted to pick up on what you said about optimizing for different demographics. When you look at the data, what are some of the notable differences that you see among different demographic groups?

ET: 04:20

Yeah, so the injury data shows that there are certain demographics that are still disproportionately being impacted, right? So you have certain crash modes where perhaps women experience higher injuries for certain parts of their bodies, or the elderly demographic is experiencing higher injury rates depending on where they're sitting in the vehicle and depending on the crash direction.

Automotive safety has come a long way and the advances that we've made have really benefited everyone. But we're at the point, and we've been at this point for a while now, where it's time to of fine tune that. And so it's not just continuing to use the same tools that we have and keep going forward and being like, OK, well, most everyone is being protected. Well, we need to really kind of hone in on the groups that are still being marginalized and really try to pinpoint like how do we address the injuries that they're seeing? How do we optimize for that so that, you we can reduce those injuries and fatalities?

ET: 05:22

We are at a point in time where we have so much technology available to us. We understand so much more about the human body. Like, let's put all that together. Let's figure out how do we solve for the differences in gender? How do we solve for the differences in age and pregnant women and obesity? And there's so much variation in the human population that we need to be trying to address that because those are the people that get into cars, right? Like there's not just one type of person that drives a car or rides in a car. And so in order to really work towards that vision zero, you have to be able to provide safety across the various body types and ages and just different demographics.

TS: 06:07

And if we look at men and women, can you give a sense of what the differences are in terms of injury rates, particularly when a crash occurs?

ET: 06:15

Yeah, so female drivers and right-fronted passengers are approximately seventeen percent more likely to be killed in a car crash than a male occupant of the same age. And data has also shown that any seatbelt-wearing female has seventy-three percent greater odds of being seriously injured in a car crash than a male in the same kind and severity of crash.

TS: 06:39

The data on injury disparities is from the National Highway Traffic Safety Administration in the US and from a 2019 study from the University of Virginia. But this isn't just a US problem. Emily Thomas says there are clear differences in the injury data globally.

ET: 06:54

When we talk with our colleagues internationally, we're all kind of saying the same things. We're all seeing differences.

TS: 07:02

Another important detail, if we zoom out and look at overall fatalities globally, more men than women are killed on our roads. Here's Astrid Linder with an explanation.

AL: 07:12

Currently, we are 1.2 million people dying on our roads yearly. So if you count numbers, if you count how many men and women are killed as drivers on our road, you will find that more men are killed than females. There are some really strong contributing factors to this that has not to do with how well we are protected, but what we are exposed to. And the male drivers are as a group, not, of course, you cannot say for one individuals to another, but as a group, they drive faster, are more often influenced by alcohol, and are less keen to wear a seatbelt than females.

TS: 08:06

As they work towards a world where there's zero fatalities and injuries, what automotive safety experts like Astrid Linder and Emily Thomas also want to see is equity in safety testing to reduce the disparities when a crash does occur. This is where crash test dummies come in.

AL: 08:23

My vision is that for 2030, of vehicles should be assessed for both females and males. We talk a lot about innovation and that's how we have created our whole society. And in this case, by not having the female part of the population represented, we are not able to identify what works best for both men and women. What it does is that you don't give those that develop the best performing innovations the credit for what they have achieved. And then they might disappear from the market without you even notice it.

ET: 09:05

So there are disparities in crash safety that we're seeing. And the tools that we have to help advance crash protection in a vehicle, we have crash test dummies. And crash test dummies historically have been designed with male data. When crash test dummies were first created, it was primarily men that were driving. That has certainly shifted. You have women that are driving just as much as men now, if not more. The culture has changed, but our tools haven't necessarily kept up with the culture. They've created female crash test dummies that are using male data and then scaling it down to a smaller size and using that as the tool to gauge female injury.

And as I was training in pediatric biomechanics, one thing that my advisor used to say all the time is that kids are not just small adults, because for the longest time, even our pediatric crash dummies were just scaled down versions of the adult dummies. So kids are not just small adults. And you can extend that to females. Women are not just small men. It's not as though we are exactly the same and you've just created a fun-sized version. And that's good enough. There are significant differences that impact our biomechanical response in a crash.

Things like the shape of our joints, the material properties of our muscles and our bones. There's various aspects of us, how our bodies react in a crash is going to be different than how a man's body does. Even like our posture, our torso, the strength of our neck compared to our heads, there's so many different facets. And if you're not accounting or you're not accurately representing those facets, then the injury responses you're getting from the crashes aren’t even giving you a full picture. So it's really important that we are trying to have better, more realistic data, and that we're not just treating women like they're small men, because that's not accurate. In order for you to have an accurate response data set to analyze, you need to have accurate inputs.

TS: 11:27

If we could just take a step back and for people who aren't familiar, could you explain the current state of crash safety assessments? How are these assessments being conducted at the moment?

ET: 11:40

So right now, crash safety is evaluated in two ways, right? So you have the regulatory testing that happens. So wherever it is that you live, vehicles in your country have to meet the regulatory standards for cars. And so there's going to be certain thresholds for vehicle safety in every place. And so typically, there's going to be dynamic crash tests that need to happen.

And there will be certain thresholds for injury that have to be met within that framework. And the dummies that are being used for that, really it's the same dummies worldwide. Because we have essentially like one major manufacturer for crash test dummies, and so everyone's using the same dummies. And so you have a 50th percentile male dummy. There is a 95th percentile male dummy. There's a fifth percentile female dummy, which is taking that 50th male, scaling it down to be a fifth percentile female. And when I say the percentiles, we're really looking at things like height and weight, right? So it's looking at for the whole population, fifth percentile height and weight is being represented. And that fifth percentile female can also be used to represent like a twelve to thirteen year old child. So it can be used in multiple applications. And then you have some pediatric dummies.

ET: 13:02

There's dummies for different impact directions. So some dummies are really designed and tuned for frontal impacts. Others are for side impacts, or they could be used for rear impacts. And so you have to have the right tools depending on which crash direction you're evaluating. So that's the regulatory testing that happens. And then there is consumer information programs, like consumer testing that happens, things like the new car assessment program, or NCAP. So there are NCAP programs throughout the world.

And these are tests that are being done above and beyond the regulatory tests, right? So in order for vehicles to pass certification and be available for sale, they have to meet the regulatory standards in your market. Then on top of that, if manufacturers want to be like, look, we're going above and beyond, we want the star ratings, we want to show that we really focus on safety, then they'll want to perform well in the NCAP program relative to their market.

Those are typically tests that are done at perhaps higher speeds and to have a higher severity crash, to tease out crash worthiness differences. Because with regulation, there's a minimum bar that's being set for safety. And so it's pass-fail. You either meet the minimum bar or you don't. Now, some manufacturers could be far exceeding the minimum bar. They're doing great. But there's no way to tell that from the regulation. It's just they passed.

ET: 14:30

Consumer testing allows for you to see how much better or worse is this vehicle compared to its peers, right? Because you're testing it at higher severities, you're teasing out crash worthiness differences. So typically, consumer testing tries to go beyond what regulatory testing is doing. And so the NCAT program is a great way of doing that. So these programs are happening worldwide, and that's kind of where we are. We're using...even though they're testing them at higher speeds or they're testing them at different constraints, you're still using the same tool, you're still using the same crash test dummies that you would in the regulatory program as well. And typically it's, you know, it's the same ones, you're still using the 50th male, you're still using the fifth female. So pretty much is being used across the globe to evaluate crash safety in vehicles.

TS: 15:18

So essentially there are variations across different countries, but the takeaway is that even though they're different types of tests, they're using the same dummy, which isn't necessarily representing everyone in the population.

ET: 15:30

Right, there's also, so there's physical crash testing. And then there's simulated crash tests that are happening with computer simulations. And so with computer models, you can make more adjustments. There's opportunity to scale them to different sizes, to adjust the material properties that are being used, things of that nature. And so there's a lot of virtual testing that's also happening. And manufacturers themselves, they do a lot of in-house testing as well before it even goes to regulation testing or before it even goes for consumer testing. And so there is a lot of work that's being done, but there's limitations to both. If you only test things virtually in a simulation, well, then you're not able to see how it's playing out in production vehicles.

You need to have an actual physical model to see, OK, well, everything looks great on the simulation. But then once you make the thing, how is it performing? If we only test in physical crash tests, one, they're incredibly costly, right? Doing a full-scale vehicle test is very expensive and you lose the car. You can only use one car per test. that's also, and it has limitations because you can only use the tools that you have and you can only...It's not as if you can go in and be like, OK, well, now today we're going to make this dummy a different size, or we're going to tweak the material properties of this dummy today and see how it reacts in a different way. You can't do that. It's what you have is what you have. So both types have limitations to them, but that doesn't mean that they're not useful. You can't learn from them. And now, really, everyone's pretty much using both types of testing to fuel their vehicle safety efforts.

And so that's how they're using it to get fuller pictures of how different body types, different sizes, different ages are gonna, how those bodies are gonna respond in a crash.

TS: 17:33

So let's talk about what it takes to create a physical dummy. What kind of data is needed?

ET: 17:38

So the crash test dummies are meant to be representative of the human body, right? But within certain limitations, a crash test dummy has to be very robust. It has to be able to withstand crash forces multiple times because we're using the same dummy to conduct multiple tests. So they can't just kind of fall apart after every test. You need a tool that can withstand multiple impacts over time, but also give you repeatable results.

So, you don't want to use the same test dummy and run a test and get one set of results, run the same test again with the same dummy, but get a completely different set of results, right? That's not helpful. And so, you know, there's a lot of research that goes into developing a dummy. You have to have biomechanical data that understands, you know, things like rib strength and things like tendon strength and how much force applied to certain different aspects. Is it going to cause a fracture? Is it going to cause strains? Is something going to tear? You're using the crash test on me to really have a simulation of what would happen to the human body. How much load can the body withstand before it gets to the point of injury? How much load can it withstand before you are now having organs that are getting lacerated or you're having bones that are breaking?

ET: 19:06

Materials are being carefully chosen to replicate. You have a lot of, crashing summons are made out of a lot of steel and rubber and all these different materials, cables to represent different ligaments and tendons and muscles and things of that nature. But it's a highly calibrated tool. And then on top of that, youhave tons of sensors that are within it to capture all of this data for various body parts. So you have sensors that are capturing loads, like the forces that are being applied that are, you have ones that are measuring the moments. like, you know, the rotations of like the neck and the spine and the joints across all the different axes and trying to determine like, what are the injury thresholds? And so you need data, like you need research data that helps from like real bodies that helps us determine like, what are the injury thresholds for different age groups, for different genders, when you're loading it from different angles. So there's a ton of background research that has to happen in order to now create a tool that's going to accurately simulate the human body. And that takes years. It's not a small feat to create a crash test dummy. There's a lot of work that has to go into it.

TS: 20:20

When Astrid Linder first had the idea to develop a female crash test dummy, one of the challenges she encountered was a lack of female data.

AL: 20:28

When we make models of the humans, of course you need to compare to the response of females. And if you have only done testing on males, then you don't have any data for females. So the first step was to ensure that there were equivalent sets of data for males and females.

TS: 20:46

She started by creating a virtual model and then a prototype of an average female dummy. Eventually she got European Union funding to develop physical dummies representing both an average female and an average male. They were designed specifically for low severity, rear impact crashes, in which studies showed the woman have a higher risk of whiplash than men.

AL: 21:07

We launched both a concept for an average female and an average male done in the same way. All the drawings are open. We work with open source and open data. So we really provided this for, yeah, for the whole community.

TS: 21:23

While Astrid Linder's work has shown that the data is available to create a female dummy, there are a number of hurdles to overcome before a new dummy can be used in regulatory testing in any country.

ET: 21:33

Even in the United States, there's the Thor dummy that's being developed. So there's a Thor 50th male, and then there's a Thor fifth female. And the fifth female is using female data for its injury thresholds and in the development and such. But there's a lot of testing that's going into getting these dummies ready for regulation in terms of like, these are dummies that are certified for use in regulatory testing. And to do that, there's a lot of testing that's happening across labs, within labs to show that repeatability and reproducibility. Because you want to make sure that regardless of where you're using the tool, that you're going to get the same outcomes, right, within a threshold window.

And then you have to get these dummies now written into the regulations for the different impact types. so, like, I mean, in the US, adding new regulations or amending regulation takes a very long time. And so there's a lot of different steps that have to happen in order for us to be able to start using it. And the regulatory bodies need to say, hey, if we're doing rear impact testing, that you're now required to test with the 50th male and the 50th female. It has to be written into the regulation. Because unless it's written into the regulation, no one's required to do it. Manufacturers might do it in their own R&D, but that's not what they're having to do for certification. And so until it's officially part of a regulation, we won't see the full value of it. There'll be some value in that it's being used for research and development. It's likely being taken into account. But this would really mandate that, you have to design for this, right? And that's we want. We want the manufacturers to be designing with women in mind. And we want it to be written into regulation so that they're having to do it. That's the best way to move forward.

TS: 23:25

So looking to the future, how do you see car safety evolving, especially with the rise of autonomous vehicles? How do you think that will impact crash testing?

ET: 23:34

There's still so much unknown for autonomous vehicles. And it's more than just having the car equipped with that kind of equipment. It's also having the infrastructure around the vehicle be ready for that. So we're not there yet. We probably won't be there for quite some time. But as we're moving in that direction, it's really driving home the idea that wherever someone is sitting within the vehicle, they should expect the same level of safety, right?

And I think as challenging as that work is to, you know, how do you take these technologies and make sure that they're going to be advantageous for anywhere from like children that are in boosters and use the vehicle seatbelt up to perhaps an elderly person in the backseat using the vehicle seatbelt. That challenge though, and really going through those like mental exercises and those engineering exercises now is going to prepare the way for making sure that there's safety equity as we move towards autonomous.

ET: 24:32

Because that's from what people keep envisioning, right? That's how it's going to be. You could sit anywhere in the car, and you could be facing in any direction, and kids won't just have to be in the back seats. They could be up in the front. It's going to change how the vehicle environment really is. And so us figuring out how do we account for various demographics in one seating position, going through that engineering challenge is going to, I believe, create the building blocks for crash safety for autonomous vehicles as well. So I think it's an important exercise for us to figure out, and that we can't delay any longer. We have to be focusing on these things.

TS: 25:14

I asked Astrid Linder and Emily Thomas what gives them hope things are heading in the right direction and what difference can we make as ordinary citizens?

AL: 25:25

I think it's always difficult in a process like this, because we are in a process and the optimism is that the process is alive. It's like any peace negotiation. I think it's always difficult in that process to see what will be the turning points. You see that afterwards. But I think all the diplomats are well aware of that, that you don't know, yeah, we have reached this point or we are close to a deal. We have unfortunately such processes ongoing in the world today. So it's easy for us to relate to that, right? We don't know, but we know that the process is alive. So that gives me optimism.

I think it's important how you vote. Depending on how you have voted, you also vote for an inclusive or exclusive or a society which excludes different parts of the population. So that's important. And that's how we form and shape our society.

ET: 26:30

I think regardless of who you are, where you are, when you purchase a vehicle, it's important to pay attention to the safety readings, right? You want to choose a vehicle that performs well in the consumer testing programs because even with imperfect tools, there's still been a lot of progress that's made. So you want to make sure you have the most safety available to you. And I always say, buckle your seatbelt regardless of where you're going, how far away it is and where you're sitting in the vehicle. Your seatbelt is your first line of defense in any crash. So making sure that you are belted so that you can really benefit the most from the vehicle crash management systems is important.

As far as what else you can do, I think it's raising awareness and reaching out to your lawmakers and to regulatory bodies and the manufacturers. mean, these are things that you guys care about, right? We want to see safety move in this direction. There is work being done. There are reasons to be optimistic about this. Just even in my interactions and attending conferences and being able to see the work that's happening, there are more studies being done to better understand seatbelt fit for pregnant drivers. Here in the United States, we have the Center for Child Injury Prevention Studies. They're doing a study that's understanding and measuring thoracic anthropometry and belt fit on adolescent and young adult females, right? Taking into account the fact that as our bodies change and, you know, bra fit and things of that nature, how does that impact seatbelt fit? How does it impact seatbelt comfort?

ET: 28:03

I think just the sheer number of like women that are coming into STEM and young researchers that are sitting at the tables, like both in academia and in industry and in government to be like, these are things you need to think about, right? You need to understand this from our perspective. I'm encouraged as I attend conferences and I'm out in industry and I just see more and more women at the table. And I'm like, that's what we need. We need to encourage our women to pursue STEM and to be speaking up about how their experience is different than the person sitting next to them as they use the world around us, right? Because that's how you create anything to be more inclusive. So there's plenty that can be done and it can even be small things like using your voice in your workplace or in your community to make sure that it is being considerate of all the different demographics that utilize it.

TS: 29:01

Thank you to Astrid Linder and Emily Thomas for a really great discussion. If you'd like to explore this topic further, I've included some resources in the show notes. And let us know what you thought about this episode. You can reach out on LinkedIn or Instagram to continue the discussion. We're at Made For Us podcast. And why not tell a friend about it as well? I'm Tosin Sulaiman. Thanks for joining me on Made For Us.