
Crash Test Science! (Hacking Physics To Save Lives)
Season 11 Episode 19 | 37m 33sVideo has Closed Captions
Crashing a car is usually a terrible accident. But sometimes, it’s a way to save lives.
Crashing a car is usually a terrible accident. But in this episode, we’ll meet four engineers who crash cars for a living. They do it to understand the physics behind a dangerous crash… so they can hack those same physics principles to protect drivers and passengers instead of endangering them.
Problems playing video? | Closed Captioning Feedback
Problems playing video? | Closed Captioning Feedback

Crash Test Science! (Hacking Physics To Save Lives)
Season 11 Episode 19 | 37m 33sVideo has Closed Captions
Crashing a car is usually a terrible accident. But in this episode, we’ll meet four engineers who crash cars for a living. They do it to understand the physics behind a dangerous crash… so they can hack those same physics principles to protect drivers and passengers instead of endangering them.
Problems playing video? | Closed Captioning Feedback
How to Watch Be Smart
Be Smart is available to stream on pbs.org and the free PBS App, available on iPhone, Apple TV, Android TV, Android smartphones, Amazon Fire TV, Amazon Fire Tablet, Roku, Samsung Smart TV, and Vizio.
Providing Support for PBS.org
Learn Moreabout PBS online sponsorship- [Both] Here we go.
(metal rumbling) (car whizzing) (car thudding) - Whoa.
- Oh my God.
- Wow.
(all laughing) Hey, smart people, Joe here.
Nobody ever dreams of getting in a serious car accident, but in the event that you do, you should be very thankful for where I'm standing right now.
I'm at the Insurance Institute for Highway Safety.
They do some of the most advanced crash testing on earth right here in this building and the proof is all around me.
Today, we're gonna do a really interesting deep dive into car crashes and show you something that most people never get a chance to see, how engineers and car designers hack physics principles that could be deadly so that they protect drivers and passengers instead.
We're also gonna learn about the intense preparation that goes into one of today's high-tech smash tests and we're also gonna see what's inside a crash test dummy and how these incredibly advanced and high-tech sensory tools have saved more lives than you could possibly imagine.
Then we're gonna look into the future and see what innovations may one day help cars avoid crashes in the first place and we're gonna smash a car.
So buckle your seat belts, this one's gonna be awesome.
(bright music) Americans take over a billion trips by car every day and every day around 18,000 of those trips will result in crashes and at around a 100 of those, someone will die.
The fact that fatal car crashes still happen every day is why car manufacturers are constantly trying to build safer cars.
But at the same time, when you think about all the driving that we do, that number is really low.
It's less than 1% of crashes.
Around a century ago, your average driver or passenger was 20 times more likely to die in a car crash and that's because in the past few decades, engineers have developed some incredibly advanced ways of studying the violent physics that happens in a car crash.
And that data has led to some amazingly high tech innovations in how cars are built, many of which you probably don't even notice.
I wanted to find out how they did it.
So I found someone who wrecks cars for a living to let me in on their secrets.
Becky.
- Hey.
- [Joe] How's it going?
- Good.
- [Joe] So you officially have the coolest job ever.
Are you responsible for everything that we see behind you?
- So I developed the next generation of car crash tests here at the Insurance Institute.
- [Joe] That sounds like a yes to me.
It seems like in this building is where reality meets Newton's laws.
Is there one thing that sticks out like one physics principle that you think drives the biggest questions in crash testing?
- Every vehicle for like their frontal crashes goes from 40 miles an hour to zero when it hits the barrier.
And so it's a question of how can we cushion that impact so that people can survive those crashes?
- That little thing Becky just said there is the key to how engineers hack physics for safety.
Every car in a crash is going to start at some speed and very quickly go to zero, but carefully controlling what happens in that moment in between, that is what determines whether people survive.
When this clicked for me, I felt like Neo when he figures out the Matrix, because I finally understood what engineers are looking for when they do this.
(car thudding) - [Speaker] Wooh!
- And so you can experience that feeling too when you just spend some time with a guy who lived hundreds of years before the first car was even invented, Mr. Isaac Newton.
Now you might think what's important in a crash is keeping the passenger compartment intact and you being strapped to the car so you don't become a human projectile.
And those are important, but there's another bit of physics that's probably even more important.
The force that your body experiences is directly dependent on how long the crash takes.
So you mentioned every car is gonna go from whatever speed it's going to zero in an instant, but does how long that instant is matter?
- A frontal crash test can be anywhere from a 100 milliseconds to 150 milliseconds.
- [Joe] Which seems like nothing.
- It's a blink of an eye.
It's super fast in terms of the world that we live in, but we can tune different parts of that to slow the vehicle down in a more controlled manner, in a more uniform manner, to provide less of a jarring experience for the occupants inside that vehicle.
- Extending how long a crash lasts is the real secret of car safety and it all comes out of Newton's laws of motion.
Newton's first law says that an object in motion will stay in motion.
It'll keep going at a constant speed in a straight line until acted on by some external force.
That's called inertia.
(upbeat music) - [Announcer] Inertia is a property of matter.
- [Joe] In a car crash, the object in motion is the car and also your body, the organs inside your body and all the stuff in the car with you.
All those things are gonna keep moving until some external force acts on them.
Now, that external force on the car could be the brakes or it could be another car or a wall.
Whatever it is, that force is going to bring the car down from its original velocity to zero, but the amount of force isn't always gonna be the same.
That's why you don't get whiplash from stopping in a red light.
We can actually figure out what that force is going to be for different kinds of crashes.
Newton's second law says that the magnitude of the force that it takes to stop an object in motion depends on two things, the object's mass and its acceleration.
If you've ever seen that equation, F = ma, that's what we're talking about.
Acceleration or deceleration in the case of a crash is just a change in velocity over time.
Let's rearrange this really fast.
When you look at this equation, you can see that there's not much to work with if you wanna make the force smaller on a car in a crash.
The car's mass is whatever it is and the change in velocity is going to be the same no matter what happens, the car is going to go to zero.
So as an engineer, you don't have any control over those, but the one thing you do have some control over is this piece right here, T, how much time the crash takes.
The shorter the time, the bigger the force and the longer the time, the smaller the force.
So say a one ton car goes from 40 miles an hour to zero in one second.
(cars thud) Let's not worry about the units right now.
The force here would be 40.
That's a really fast deceleration and a really big force.
If that same car going the same velocity stops over 10 seconds, that's a much slower deceleration.
The force for that one would be four.
The car experiences 10 times less force with the longer deceleration.
This is why Becky says the key to making crashes less dangerous is extending the time of the crash in a carefully engineered way.
Sometimes by just milliseconds because any slowdown helps reduce the force on the passenger compartment and that can make the difference between a dangerous crash and one that you walk away from.
Let me show you what I mean.
Okay, I've got two eggs here and I'm gonna sacrifice at least one of them in the name of science.
I'm gonna throw one of these against a hard surface and one of them against a sheet.
Now both of those things are gonna bring the eggs' velocity down to zero, but the sheet's got a little bit of give in it, so it should bring the velocity down to zero a little slower.
Let's see what that means for the eggs.
(egg thuds on wall) (Joe laughs) (egg thuds on sheet) The physics of how to break an egg isn't that different from the physics of how to survive a car crash.
So extending the time of deceleration makes a crash safer.
If all of the front of this car crumpled up in 50 milliseconds, then that would be more dangerous than all of this same exact deformation happening in a hundred or 150 milliseconds.
- Exactly, the same amount of deformation happening in a much shorter time period is much more dangerous for the people riding inside that vehicle.
- More than anything else, this concept is the key to building safer cars and car manufacturers have come up with all kinds of ways to exploit this physics trick in their car designs.
Weirdly enough, they do that by building cars that are designed to get wrecked.
Now, in the olden days, cars were built to be sturdy, which that seems intuitive, right?
If I were building a metal capsule to carry people around at highway speeds, I'd want that thing to be pretty strong.
But decades ago, car manufacturers realized that building cars whose ends would crumple would extend the time of deceleration in a crash just a little, and that would reduce the amount of force on the car as a whole.
That meant that in any serious crash, the car would likely be destroyed, but hopefully the people inside wouldn't be.
I know extending a crash by a few milliseconds doesn't sound like much, but I've got a little experiment here that can give us an idea how a little bit of a cushion can translate to a big difference in force.
This looks like a toy car, but it's got some sensors inside that can measure how many g's of deceleration are experienced at impact.
If I run this car into the wall right now, it's gonna stop very rapidly, maybe even bounce back just like those sturdy cars built before the 1950s, and since it's gonna stop really quickly, there's gonna be a lot of force on this car.
(toy car thudding) But if I give it a little bit of cushion at the front, that will act a lot like a crumple zone would.
(toy car thudding) Let's compare the deceleration data that the car recorded with and without the crash cushion.
Without the crumple zone, the car experienced nearly 16 g's of deceleration in less than a hundredth of a second.
It even rebounded backwards causing another eight or nine g's of jarring acceleration in the other direction.
But with a crumple zone, the car experienced a max of only six g's of deceleration.
What made the difference was this, that the deceleration and force was spread out over this much longer time period.
Crumple zones have been around since the 1950s because they work.
This is a car that Becky's team crashed a while back.
And under the hood here, you can see part of a big structural beam that runs lengthwise down the front of the car.
It's one of two frame rails that are meant to let the cars front end crumple in a slower controlled way during a head-on collision.
Every car today has these, because of what we learned about Newton's second law, this innovation saved tons of lives.
But engineers noticed that some people were still getting injured or killed in frontal collisions because not every collision is gonna be totally head-on.
Car crashes happen from all kinds of different angles.
These days, some of the most dangerous accidents are collisions that only hit part of a car's front that miss the frame rails entirely.
And so crash test engineers developed a new crash test to push car manufacturers to solve that problem.
- This is a different one of our four high-speed crash tests that we run.
This is a small overlap, so only about 25% of the vehicle front end strikes our barrier, and this represents a vehicle striking another vehicle or a tree, post, pole, those kinds of crashes where only about a quarter of the vehicle interacts with an object.
It kind of bypasses a lot of the main structures of the car.
This main frame rail is often used in our moderate overlap crashes to absorb energy.
In this case, it comes right past that.
So we need to look at different structures further outboard on this vehicle, including the wheel and the whole assembly of the wheel that can be used to absorb that energy.
- [Joe] Okay, so I see that those things in the other vehicle we looked at that were extending the time of that deceleration, they're not getting engaged in this crash.
- At all.
Yep.
- [Joe] So you have to engineer new structures because like the suspension has another job, like it's supposed to do its thing.
The wheel has another job, brakes have another job, but they're almost pulling double duty.
- Yeah.
And prior to us introducing this test about 10 years ago, all of these components were just doing the job that they were assigned to do.
The wheel helps you go forward, the suspension helps, you know, stabilize the vehicle.
Now all of these structures are having to be re-imagined.
- So, instead of just relying on the frame rails, engineers are now designing everything in the front of the car to crumple, including things like the wheels and the suspension.
That means that a car that looks really bad after a crash like this one, did exactly what it was supposed to.
It slowed down the crash by letting a bunch of different pieces crumple and extend the time of collision by milliseconds.
This is a really cool case of safety testing, inspiring engineers to solve problems, and then creating harder tests that highlight new problems for engineers to solve.
It's a feedback loop that saves lives.
But designing safe cars isn't just about the cars, it's also about how the humans move around inside those cars or hopefully don't move.
That's why seat belts are a thing.
So one thing I remember is an object that's in motion stays in motion until it's acted upon by an external force.
- Exactly.
- [Joe] The wall in a crash test is the external force.
- Yep.
- [Joe] And the object in motion is not just the car, but the person inside, right?
- Yep.
You inside the car, and we wanna make sure that the person inside the car stays coupled to the car and doesn't strike the things within the car as the car is slowing down, but the person is not.
And that's really where seat belts and airbags come into play.
Engineers can figure out how to decelerate the car and design the car so that it's survivable, but it needs you to be coupled to that car in order to provide the most protection.
- [Joe] So the seat belt's tying you to the car so that the engineering can work.
- Exactly.
- This was also a really counterintuitive thing for me to figure out.
I mean, seat belts do keep you from becoming a human flying projectile, but their most important job is making you and the car act as one physical body so that you experience the same extended deceleration that the car does.
But that's kind of a double-edged sword because g-forces in a crash can be huge more than what astronauts and fighter pilots feel.
So seat belts have a second job reducing the forces that you feel, and they do that in some insanely cool ways.
In some cars right at the moment of impact, the seatbelt tightens up a little bit.
This is called a crash tensioner.
You might have felt this if you've been in a car when the driver slammed on the brakes.
- Hey, I'm walking here, I'm walking here.
- That's to help tie you to the car so that you decelerate along with it.
But some seat belts go even further.
Some are now engineered not just to tighten up and lock, they'll also unspool a little during a crash just to slow your body down for a few more fractions of a second before it comes to a stop.
- Another way we can really extend the time of deceleration for the human inside is to put things like airbags in the vehicle.
So these are air cushions, essentially.
They're pyrotechnic devices.
They inflate during the crash and they allow the dummy to kind of ride down parts of the crash in ways that...
Almost like a catcher's mitt, you're kind of catching and holding the dummy so he can't contact hard structures and he's got less forces.
And so in many of our frontal tests, the side airbags also deploy to keep the occupant contained, but it's really that frontal airbag that can be used to improve and extend the time.
- [Joe] You have like a perfect face print.
- There's a perfect face print on this one.
- [Joe] I always intuitively thought of airbags as a shock absorber as like a spring.
Like it's just, it's taking my force away, but wasn't as intuitive to think of it as extending the time, again, that I'm decelerating.
That's really cool.
- Each of these cars is designed with over a hundred different safety features that are meant to provide little protections here and there.
Even the materials on the dash, the shape of the dash, there may be energy absorbing components here that when your knees go into the dash, it's not as injurious as if it was just a metal plate sitting down there.
- [Joe] Every part of a vehicle is engineered- - Every part.
- [Joe] With these things in mind.
- They shoot dummy heads into different parts of the car to make sure that it's safe, even if you're not wearing your seatbelt, that they're... - [Joe] They have a dummy head cannon that they- - They have a...
It's basically a dummy head cannon and they shoot it in all of the upper structures of the vehicle.
So these plastic parts and fabric parts of the car have to still provide a certain level of head protection.
- [Joe] So it's not just comfort.
Everything is engineered with safety in mind.
- Anything that's in this occupant compartment is gonna have safety designed into it.
- So engineers are constantly hacking Newton's laws to make cars safer and safer.
That's really cool and it's totally changed the way I look at how cars are designed.
But seeing how all this will play out in real life is pretty hard to predict because there's just so many variables in any car accident, and that's a problem only a dummy can solve.
Crash testing started in the 1930s.
It was a way to see what parts got damaged and how different safety features worked.
Now, these days, engineers do a lot of this stuff in simulations, but there's still no replacement for smashing a car in real life and just seeing what happens.
This is Sean O'Malley and he is preparing this car for one of those super challenging tests that Becky mentioned, where only the outside of the car strikes the barrier where those frame rails aren't much help.
And no test is complete without a roll of duct tape, it looks like.
- If you look at this vehicle, probably a roll of duct tape on it right now.
(Joe chuckles) The yellow and black tape is used to determine the center line of the vehicle.
These stickers are a set distance apart, 75 centimeters between them.
And on the doors, they're 50 centimeters apart.
All of these distances are in our test protocol.
Any researcher afterwards could look at our crash videos and do their own video analysis of crushes dummy movement and using just the pixels- - [Joe] A known distance.
- That's a known distance.
And there's three or four of those known distances on this vehicle.
- [Joe] And so you said you were aligning this vehicle for tomorrow's test, so it's gonna hit this thing behind you?
- Yeah, this test is a 25% of the vehicle width overlap into a rigid barrier.
Between these two lines right here is exactly 25% of the width of the vehicle.
Our test protocol dates plus or minus 1%.
I'm in the process of putting this target on so we can visually see during the video analysis afterwards of where it actually hits.
- [Joe] Okay, so that'd be perfectly- - [Sean] That's center.
- [Joe] Aligned with the edge of that.
- And looking at the video afterwards, as long as this corner of this target is between here it meets the protocol.
- [Joe] Then the test was a good test.
- The test was a good test.
- [Joe] Okay.
Turns out there's way more to a crash test than just slamming a car into a wall.
The team actually spends days getting the car ready.
- All internal combustion engine vehicles, they're sent to the VRC.
We receive them, we take photographs of them and then before we prepare them, Joseph measures the structure pretest about 12 locations then we send the vehicle to the vehicle prep guys who drain the gas from it and replace it with Stoddard, which is a mineral spirit.
They drain coolant, engine oil, transmission fluid, all the oils that are gonna make a mess after a crash.
We set the seat for the dummy, then we roll it out here to the crash hall, set it in front of the barrier.
- [Joe] So, yeah, we're getting tomorrow's star of the show into the driver's seat.
- [Sean] He is the star.
- And he really is the star.
In any car crash, there's obviously the collision between the car and whatever it hits, but there's also the collision between the body and the car and the collision between a person's organs and their own body.
So reducing damage isn't just about physics and engineering, it's also about biomechanics, understanding the effect of physics on human bodies.
To understand this piece of the problem, engineers used to stick cadavers in cars during crash tests, but that was kind of weird and not really ideal for anyone.
So pretty quickly they found a better way to do things.
Jessica Jermakian does research on how people get injured in crashes and how to make crashes safer for us human sacks of meat by using mechanical human surrogates.
Crash test dummies.
- The first crash test dummies that I'm aware of were from back in the 1940s, came out of the Air Force when they were testing ejection seats for pilots.
And so a lot of the early work was done on military men, including many volunteers.
But the automotive industry and even the military realized that this was a really great tool for looking at occupant protection.
That military ejection seat dummy has morphed into the crash test dummies that we see today.
- [Joe] This is called the Hybrid III.
This type of dummy is the workhorse of crash testing.
It's been in service since the 1970s.
- If you look at this dummy, you can see that it's a fairly complicated tool, but it is not so complex as the human body.
It doesn't have, you know, bones like our bones and you know, organs like our organs.
It's made of a bunch of metal and vinyl and some sensors.
- [Joe] Not a lot going on upstairs.
- Yeah, not a lot going on in here, but we do have a lot going on in here.
In that we can put sensors inside the head.
So we have an accelerometer here.
We have a load cell where we're measuring forces in the neck, forces and moments.
We can measure the compression of the chest.
This is tied to a sensor down here that is measuring how much that moves when it is pushed inward from things like the seatbelt or- - [Joe] So the more that moves, you're just, you're getting a stronger signal.
- Yes.
- [Joe] Based on pressure.
- Yes.
Based on that information, we can back out how many millimeters of chest deflection.
And we know from historic data with things like human volunteers or cadavers or animal studies, we can translate those deflection measurements into injury risk in real people.
- [Joe] To put it one way, maybe, from metal to meat.
(Joe speaks indistinctly) (Jessica laughs) So this generation of dummy right here has saved more lives than we can possibly imagine through the advancements that it's led to.
Not you literally.
(Jessica laughs) The dummy getting ready for this crash test is tricked out with all kinds of sensors and all of them feed data into these boxes in the back of the car.
But not every aspect of the crash test is so high tech.
- Right now, Tyler's putting paint on the dummy on the head and the knees, shin and the left hand.
It's basically grease paint, which will show us after the crash, the impact locations.
- [Joe] Wherever that stuff touches, it's gonna leave a mark.
- Yeah, I mean the airbag will have a nice face print on it ideally, and we'll know, you know what... Based on the color, what part of the head hit the airbag.
- [Joe] It's like a Rorschach test for you.
- Yeah, some of 'em are actually pretty attractive.
- [Joe] So if you've mastered finger painting in kindergarten, then you are- - You're hired, you're hired.
- [Joe] This is the job for you.
Okay.
There's still a lot of prep work going on in and around the car, but the big lifting is gonna happen underground.
Sean brought me down to show me the machinery behind the crash test.
- Nobody's allowed...
I said nobody's allowed down here, typically.
- [Joe] Oh, look, I love it.
Absolutely no entry except for- - Yeah.
- [Joe] Me following you down here.
(chuckles) - The concrete barrier the vehicle will run into is right there or above us.
This is the crash machine.
It's a little hydraulic pump fired up.
It pressurizes all these accumulators down this side over here.
- [Joe] There's a button that releases the pressure from all these pumps at once into a hydraulic motor.
- This hydraulic motor right here, it's a hydraulic motor, it's runway A right here.
That's the one we use for frontal crash tests.
And there's a cable that runs all the way up and down to the shed where the vehicle is, loops around and comes back and it just pulls the vehicle in.
- [Joe] You can't teach that dummy to just hit the gas pedal.
- No, we cannot.
Crash happens right here.
- Nobody else gets down here.
This is special access.
(Joe chuckles) Back upstairs the car is almost ready to go.
The dummy is all made up and he's got all his sensors connected.
All right, so what's the next thing that's gonna happen to our friend?
- He will be launched into the airbag.
- [Joe] Have the most interesting 14 seconds of his life.
- Yeah.
- [Joe] Today.
- He'll drive 600 feet up until 40 miles an hour and that's the end of that.
(machine whizzing) (car thuds) - Whoa.
- Oh my god.
- Wow.
(all laughing) (car thudding) - All right, so now we're heading down to the vehicle to take a closer look.
- [Joe] That was so cool and I cannot wait to find out if the engineering worked.
(chuckles) - It's always really exciting to get down closer to the scene and really investigate what's happening.
So the first area that we wanna look at is this front area.
And this is an area where we don't care how much damage there is.
Actually probably more is better because we want energy to be absorbed in this front part of the vehicle.
What we don't want is any kind of destruction in the occupant compartment, the safety cage, the area where people are sitting.
So up here we've got, you know, the mainframe rail itself has barely been touched.
- [Joe] Now remember that in more head-on impacts, that mainframe rail is the main way of absorbing energy and increasing the time of deceleration.
In this test, Becky and her team are looking to see how other structures in the car absorb that energy instead.
- But 25% overlap on this car comes right outside of that frame rail.
It just skirts the very edge of it, which means that it can't do the job of absorbing all that energy.
We need to rely on other structures in the car.
And what manufacturers typically do is look at these upper structures.
The wheel itself can be used to absorb some energy, so that's why- - [Joe] It's destroyed.
(chuckles) - It's just shattered.
So a shattering at it really absorbs a whole chunk of the energy that's occurring.
And then also strengthening this area here where we transition from that front of the vehicle to the occupant compartment.
And this is the area where behind it you can imagine the dummy's feet are sitting so it's very important- - [Joe] I feel like that's where the pedals are, right there.
- The pedals are right there.
That footrest is right there.
We also don't wanna see this wheel being driven back into that zone and potentially like penetrating into the cabin of the vehicle.
- [Joe] Yeah, that's a large potentially deadly metal projectile.
- Yeah, there are definitely some cases that we've seen where that happened, it was not good.
Oh yeah, this is crazy.
- [Joe] This is a headlight lens that should be up here that is now back here.
(chuckles) That just shows you the incredible force that's at play in something like this.
Yeah, I came in here thinking damage is bad.
Like a safe car should be a car that doesn't look like it was in a crash.
This is telling me that the right kind of damage in the right place is what you're going for.
- Right, it's really essential that we be able to absorb that energy so the energy doesn't get absorbed by you as the person riding in the car.
The front of the vehicle is an area where we don't care what happens.
This car's gonna be totaled no matter what.
So let's absorb all the energy up there.
Let's destroy as much as possible so that we keep this area where the people are sitting intact.
- [Joe] Speaking of which, let's go see how that driver's doing.
- We turn off the airbags to this side of the vehicle because we wanna be able to use our cameras to really get a good view of what's happening to the driver.
The paint that we saw go on earlier for the dummy, we can now see where that paint has transferred within the vehicle.
We wanna make sure that it's only on things like the airbags, but if we do see something on the the A-pillar, the steering wheel, the dash, that would be something that we would investigate further to find out whether the dummy struck those parts and whether there would be a risk of injury because of it.
Between looking at the paint evidence of reviewing the high-speed video footage, we can get a good idea of how well the airbags and the seat belts performed to protect the dummy.
We'll also shortly be getting information from our control tower about the dummy injury measures that give us another picture of what areas there may be higher risk of injury recorded.
- They have got a ton of work to do to really give this car a grade and figure out what happened here.
But all I can say was that was awesome.
This is the coolest job ever.
And apologies to the dummy.
Thank you, Becky.
Crash tests like this have helped cars come a long way over the last few decades.
The data from these crashes gets shared with car manufacturers to help them invent new vehicle technology.
And over the last 50 years or so, car crash deaths have been dropping pretty steadily.
To see just how far cars have come in the last 50 years, check out this crash test between a 1959 Chevy and a 2009 Chevy.
(machine whizzing) (cars thudding) I know which one of those cars I'd rather be driving.
This is how smart engineering saves lives.
The National Highway Transportation Safety Administration estimates that in the 50 years between the release of those two cars, half a million lives were saved in the US thanks to innovations like the ones that we've seen.
As useful as crash tests are though on another end of the facility, engineers are working on making these tests as irrelevant as they can by preventing crashes in the first place.
What is your position here?
- I'm the vice president of active safety here at the Insurance Institute.
- [Joe] Okay, so we know about crash tests, but the ideal scenario is that you don't have to get in a crash.
- Yeah.
Obviously, you know, you don't wanna be in a crash and so if you can have a system that avoids the crash altogether, no injuries, that's the the perfect scenario.
- [Joe] So you've seen technology in vehicles just take huge leaps, I imagine, at that time?
- Yeah, I mean we went from crash avoidance not even being a thing to now it's, you know, sort of the next frontier and it's sort of standard on most of these vehicles.
- [Joe] So you're standing next to a friend.
- Yeah.
- [Joe] Can you tell us a little bit about your buddy?
(Joe chuckles) - Yeah.
So this is our pedestrian mannequin.
This is what we use to conduct our pedestrian test.
He rides on sort of a surfboard in a belt.
The system has GPS that's constantly coordinating with the vehicle for the position.
His legs actually move just like the pedestrian would be walking.
The nice thing about him is no matter what, he continues to walk out in front of the vehicle (indistinct) - [Joe] He never learns his lesson.
- Yes, he doesn't.
So we have behind me a Genesis GV80 and he's gonna go back to the start.
He's gonna come up to speed at about 25 miles an hour.
Our driver is gonna just maintain that speed.
The vehicle's gonna warn him, it's gonna be flashlights, he's gonna ignore that.
And then hopefully if all goes well, the vehicle's gonna automatically apply the brakes to avoid a collision with this pedestrian mannequin.
- So how do the vehicles figure that out?
Modern cars are designed to use a combination of technologies to detect potential crashes before they happen.
This is cutting edge technology, but most of the sensors are based on pretty straightforward physics.
That one looks really weird in front of you.
It looks like big robot eyes.
- Yeah, this is almost like a robot or cyborg and you can see it has two cameras and this is Subaru's EyeSight system and they call it that because it has two cameras that work just like our eyes.
And so they're able to triangulate the position and give the information to the computer to determine what the vehicle needs to do.
- [Joe] So the same way that our eyes using binocular vision, parallax, things like that, that computer's making those same decisions by comparing the two images from those two cameras.
- Yeah, absolutely.
And you know, one thing that's pretty interesting too is a lot of vehicles now only have a single mono camera, but this camera can still use information, determine the distance.
So for instance, in the US all license plates are the same size and so it can calculate based on how big or small that license plate is, how far the vehicle is away.
And so, you know, you can do a lot of the same things with a single mono camera.
- Some cars also use LIDAR.
They basically bounce light off of things to figure out how far away they are.
- Yeah, so this is actually a LIDAR.
So this uses light, it sends out a signal and then it receives those and it basically is calculating the time of flight.
So you know, it knows when it sent the signal out and it can calculate how long it takes to come back using, you know, the speed of light and it can detect or determine the distance the obstacle is in front of the vehicle.
- [Joe] This one looks super space age for- - Yeah, so this is a more advanced LIDAR.
So this is still using light and actually there's a spinning mirror in there.
And so this gets you basically a 360 degree view of everything around you.
Very similar to maybe what you've seen on top of the Google car or other sort of more autonomous vehicles.
And it's sort of painting a picture with dots of what's around the vehicle.
- That's really cool.
Some other tools are using sound to detect obstacles instead of light.
This one looks familiar.
I think I've...
These look like those things that are all over my bumper.
- Yeah, so these are the ultrasonic sensors.
You'll see little round ones, usually four or five on the front and four or five on the back.
And again, these are these sensors that are beeping at you when you get, you know, close to your garage as you pull in or your mailbox if maybe you're not backing out like you should be.
And you know, these are for much lower speeds and shorter distances, but again, can really reduce costs associated with low speed insurance claims.
- [Joe] So these are using sound waves.
- Yep, absolutely.
- [Joe] Versus light waves, which for some of these other sensors that must feed into how effective they are in various scenarios.
- Yeah.
And so, you know, these sensors wouldn't work well for higher speeds because of their limited range.
But you're right, it's using a sound wave and so it's like a bat, you know, around the edge of your car that's determining where different obstacles are.
- I'm just thinking now bats listening to cars backing up must sound like screaming going on.
- Yeah, who knows what would happen if there was a bat behind you as you backed up, yeah.
- So as a car's driving down the road, it's putting together information from all of these different devices to paint a picture of what's happening around it and figure out if it needs to warn the driver or even hit the brakes automatically.
All right, well, I wanna see what happens to our friend here.
We'll step out to our safe distance and then we'll...
Fingers crossed for you.
All right, so he's down at the end there.
He's gonna come here at 25 miles an hour.
- [David] Yep.
He's gonna maintain a constant speed of 25 miles an hour.
You'll see once he reaches a certain point that will trigger the pedestrian to start walking in front with the legs moving.
- Getting close.
Getting closer.
(chuckles) What?
Okay.
He survived.
He survived.
Our friend lived.
I'm happy for you.
Look both ways next time, seriously.
But pedestrians aren't the only obstacles you might need to avoid on the road.
So engineers are training cars to recognize other cars, people on motorcycles, even large animals.
And that's just one example of how improving car safety is never over.
People like David are working on making cars better at recognizing and stopping potential crashes before they happen.
People like Becky and Sean are testing different kinds of crashes so that car manufacturers can keep their passengers safe in any kind of accident.
And people like Jessica are making crash test dummies more and more human-like so that they reflect what truly happens to our bodies in a crash.
What's incredible about preventing death and injury in car crashes is that at first it all sounds like physics and at the most basic level it is, but it's also about bodies and behavior.
And solving car crashes is about combining what we know about all of those things to make cars and roads safer.
- Science and Nature
A series about fails in history that have resulted in major discoveries and inventions.
Support for PBS provided by: