Kind: captions
Language: en
[Music]
On a stretch of tarmac on the west coast
of Scotland.
Team JLB confirm all cars are good to
go. Please. For the first time ever, a
team of automotive experts will attempt
to create something extraordinary.
We have a runoff. A wildly ambitious
experiment to find out how drivers
react.
Rolling, rolling in the F-150.
And vehicle safety systems respond.
There is no take two. Real life
multi-vehicle crash test. What the
hell's that? Oh my god.
But how do you design such an
experiment?
We can't put people in the cars because
we're creating a massive collision yet
keep everyone safe.
I've just crashed.
Frankly, we don't know what's going to
happen.
Ultimate Crash
Test right now on
[Music]
NOVA. As an American-based supplier to
the construction industry, Carlilele is
committed to developing a diverse
workplace that supports our employees
advancement into the next generation of
leaders from the manufacturing floor to
the front office. Learn more at
[Music]
carile.com. Multi-vehicle pileups can be
deadly.
Each of these highway crashes is unique
with different causes and different
outcomes. But analyzing these events is
difficult. Forensic crash investigation
can only study the aftermath of these
deadly incidents by picking through the
debris left
behind. They rarely get to see exactly
how the crash unfolded.
But what if there were a way to
scientifically study something this
unpredictable, complex, and dangerous?
To stage a real life high-speed pileup
without putting anyone in harm's
way. Such an experiment could produce an
unprecedented amount of data on cars and
their
drivers. Data that could be used to
focus further automotive safety research
and ultimately make safer cars and
better drivers.
An experiment that would put crash
analysis of a major pileup to the test
to discover just how accurate it really
is. It would be a huge engineering
challenge, but one team believes it can
create a real life high-speed pileup.
It's never been done before. It's full
of technical challenges that we have to
overcome, but there will definitely be
something to learn from the event as it
unfolds.
Making cars safer is a challenge
engineers have been wrestling with since
the earliest days of car design. Your
vehicle's job is not just to get you
from point A to point B. It's to get you
from point A to point B safely.
Car manufacturers strive to develop ever
safer vehicles, spending billions
designing cars that can perform better
in laboratory crash
tests. It's a practice that was
introduced by General Motors almost a
century
ago. They were the first to do the very
basic first crash test back in 1934.
This was really the curiosity of the GM
engineers at the time to understand
their product and start to improve
safety. Nowadays, organizations like the
Insurance Institute for Highway Safety
in Virginia have taken testing to the
next
[Music]
level. Here within this temple of
science, cars are destroyed by remote
control. week in week
out, all in a quest for better road
safety. The work that we do here is a
lot of fun. There's a lot of
opportunities to crash vehicles, destroy
stuff, but it has real world meaning and
everything comes back to that real world
element.
The vehicle is going to come through
those doors, enter the crash hall at 40
miles an hour, and strike with 40% of
the vehicle's front end, the driver's
side, striking this part of the
barrier. Before the crash, technicians
apply a special grease paint to the
crash test dummies and will transfer to
anything they hit during the impact.
If this dummy hits the airbag, we're
able to really see what side of the face
as well as what part of the face or the
legs are hitting parts of the car. Could
I have an accent trigger check? You
should have it. Thank you.
The crash car itself will have no
driver, but will be propelled along a
track so the vehicle can maintain a
constant speed and direction.
[Music]
Charging is complete. 10 minutes in 3 2
1 zero.
[Music]
So, the dummies give us a summary sheet
of their results and what uh injury
metrics they consider uh would be a high
risk of injuries. And all of these
indications show that both the driver
and the rear passenger are looking good.
Survivable crash, low risk of injury.
The data collected from single car
laboratory crash tests help inform ever
safer vehicle
design. But how does
this compare to
this? So the beauty of a lab-based crash
test is everything is controlled.
Everything is repeatable. But
unfortunately, the real world is not a
laboratory and there are many, many
variables. So, I think that's the
limitation of the lab. It's almost too
good. Which is where James Brighton
comes in. He's head of the advanced
vehicle engineering center at Cranfield
University, 50 mi north of
London. He wants to run an experiment
that will combine the crash lab with the
real world by adding real drivers to a
real crash.
To achieve this, James needs to come up
with a plan that not only creates a
realistic pileup, but also doesn't put
anyone in
danger. This idea of of a multi-vehicle
collision isn't new. However, the scale
of this is really very large. We need
some very advanced systems that are
going to work to create the collision,
but also we need some very advanced
systems to be able to capture it.
This experiment has several
aims. It will investigate how different
cars fare in a multi-vehicle high-speed
crash. Analyze how ordinary drivers
react when in such a nightmare
situation and put forensic crash
investigation to the test to see how
accurate current accident analysis
really is.
All with the objective of improving our
knowledge and understanding of multi-car
pilots and making roads safer for all of
us.
But can James and his team rise to the
challenge?
The automobile has created a lot of
freedom, but we have created this
machine which kills people and we have a
duty to try and reduce its impact and
improve safety now and in the future.
A multi-vehicle pilot needs multiple
vehicles. So, James' first job is to
select the cars he wants to
crash. Eight cars will take part in the
pileup, each with a different size,
weight, wheelbase, and engine
layout. Selected to match the randomness
typical of a realworld pileup.
A trailblazer of the hybrid era, the
2012 Toyota Prius. A sporty convertible,
the 2002 Porsche Boxster Series 1. A
high-performance hatchback, a Volkswagen
Golf GTI from
2009. A pickup truck, the 2012 Ford
F-150, a luxury sedan, the 2005 Audi
A8. a 2010 minivan, the seven seat Dodge
Grand
Caravan, a sport utility vehicle, the
2005 Mercedes
ML, and finally a commercial van. In
this case, a British 2012 Vauxil
Vivaro. The cars all have very
interesting characteristics in terms of
their size and their mass. And so we
want to try and represent a nice broad
scope of cars that would be
representative of modern hyena
environment. Four of the cars will be
driven by experts from James' team,
Mark,
Aisha,
Pete, and
Chunwi. The other four drivers are to be
specially selected from the great
motoring public.
I've had um a few accidents um over the
past few years. I would say I'm behind
the wheel seven days a week easily.
People that can't drive or shouldn't be
driving. They irritate me like the most.
I don't think there should be speed
limits on M25. I'll get you to A to Z
real quick.
After an intensive selection process
examining driving style, experience, and
psychological suitability, the team
chooses four drivers.
First up is American surfer 57year-old
Tito.
The reason why I love surfing is it
allows me to escape from the
hustlebustle from from from the city
life. I come to the coast and just
escape from it all and just be one with
the sea. That's what I love.
Having driven in both the UK and the US,
Tito believes he's developed into a
confident yet chilled out driver.
I'm not a perfect driver, but I do abide
by the rules. And I'm not a wild driver.
I have a car that doesn't go fast. It's
just a
cruiser. The next driver is the least
experienced road user.
My name's Caitlyn. I live in Liverpool
and I'm 19. I've been driving now for
around 6 to 7
months. In the US, there are about 25
million drivers aged between 17 and 24,
roughly 10% of the
total. And this age group is involved in
more fatal accidents than any other.
With only a few more years experience on
the road is Luke, a 26-year-old
construction worker.
Bang bang.
As a driver, I'm more of a boy racer. I
love driving really fast when it's safe
to do so. But at the same time, you
know, there's got to be an element of
risk to really get your adrenaline
pumping. In the US in
2022, young male drivers were more than
twice as likely to be involved in a
fatal accident than young female
drivers. The final driver is 66-year-old
care worker and grandmother Lynn. Make
right. Right.
It's always a bloody man. In the US, 22%
of drivers are aged 65 and over.
I'm quite a confident driver. I'd rate
myself
maybe seven, eight. Eight. Yeah. Push
into
nine. Possibly a 10
[Laughter]
sometimes. To make the pileup as
realistic as possible and to capture
genuine reactions, drivers must not know
the experiment involves a multi-car
collision. They have just been told its
aim is to improve road
safety. The selection process included a
psychological evaluation to ensure they
can handle the pressure and emotional
challenge of the experiment to
come. Observing these four drivers
throughout the experiment will be
Natasha Merritt, an experimental
psychologist specializing in human
behavior on the
[Music]
roads. She'll be analyzing the human
factors
assessing how the drivers react when
faced with an unavoidable high-speed
pileup.
This is a bit of a hybrid between real
world and a more controlled environment.
So, uh you're seeing drivers actually
controlling the vehicles. That's really
unique, I think. With cars and drivers
selected, the experiment now needs a
location.
A nearly two-mile long airport taxiway
at a former military base on the west
coast of Scotland is the perfect
setting. It's wide enough that we can
create our slow lane, middle lane, and
fast lane, plus a hard shoulder, and a
lane for oncoming traffic.
With the cars, location, and drivers
locked in, there's one outstanding
issue. Pileups can kill.
So, how can James Brighton guarantee the
driver's safety?
So, essentially, we can't put people in
the cars because we're creating a
massive collision, which sounds obvious,
but that's that's our starting point.
James' plan to keep the driver safe is
to adapt his fleet of cars so that each
one can be driven remotely.
Eight small city cars will be
transformed into control pods, sending
signals from a safe distance to the cars
on the
track. These signals will trigger
actions like turning, accelerating, and
braking. But for this to work, cameras,
radio links, and robotics need to be
fitted. Pneumatic pistons will operate
the brakes, while electric motors will
control the accelerator and steering.
release. That's it. Perfect. Very good.
So, the person in the pod will press the
brake pedal. That will send a signal to
this car, which then makes this actuator
move to push that brake pedal to the
same amount the occupant in the pod is
pressing the
pedal, which means the car responds
perfectly to the driver's actions,
however subtle.
So, we've got steering, brakes, and
throttle all done on this one. Yep. And
same on the Golf. Yep. Brilliant.
They now need to transform the fleet of
compact two-seater smart cars into
control
[Music]
pods. So, we need to tune the pods so
that the behavior of the real car feels
realistic to the driver. So, for
example, when they move the throttle a
certain amount in the smart car, we want
that to be proportionate to the
acceleration you would expect if you
were sat in the real car at that moment
in time.
For the drivers, the pods will look and
feel like regular cars. However, their
actions won't control the car they're
sitting in. Instead, they will send
commands to the specially adapted track
cars. So, this is an encoder to measure
steering wheel position.
We've actually got a sensor that's got a
lot of resolution. So, we can measure to
07 of a degree, which is quite important
when you're if you imagine driving on a
motorway. The amount of steering input
you put to change lane is is quite
small.
So, Matt, would you like to go and press
your foot on the brake pedal?
Anything? Yeah. Press the throttle.
There we go. With the equipment fitted
in the control pods, James now needs to
get those pods to talk to the test cars.
So, we need a radio network that will
take the signals from the pods and then
send a signal to the cars on the other
end of the runway.
James has chosen to use a point-to-point
radio network similar to a walkie-talkie
system. Once on location in Scotland,
the stationary control pods housed off
the track in an enclosure will send data
to the track cars via a single radio
relay. And so that the drivers in the
pods can see where their cars are
going, the eight track cars will send
the live camera feed back to the control
pods.
Will it work? It will. It will work. Of
course it will work.
Okay, Kelvin, can you go full lock
left and full lock
right now? Really quickly like
that. Perfect.
The Ford pickup has been successfully
paired with its control
pod. Now it's time to see if it works on
the test track at Cranfield.
[Music]
Exciting, isn't it? It's good, isn't it?
Yeah. A video screen is positioned in
front of professional stunt driver Paul,
who today will operate the remote
controls. The screen will display the
live feed from a camera placed at eye
level in the driver's seat of the moving
pickup.
A little bit down. Just a tad.
That's it. Perfect.
[Music]
Okay, trackers clear and live. Okay,
we're in first gear. So, moving off in 3
2 1
go. It's looking good.
So, when you get to the end, Paul, turn
tightly and come down the middle. Yeah.
Okay.
And then basically at this speed just go
left and right. So turning right a bit.
Yeah.
Left a bit.
Yeah. It feels good. The hard thing to
judge is depth. The robotics are
performing well.
So what do you think then guys? We just
need a bit of work to do on the brakes.
That's all. I think it's a little bit
rough around the edges and the speed
we've been doing. We got to more than
double that.
Part of the experiment is to study how
the cars themselves fare in a
crash. Blackbox recorders will provide
huge amounts of
data. And helping James evaluate how the
cars and their safety systems perform is
crash analysis expert Janet Bahuth.
So my main interest will be looking at
the damage of the vehicle and to see how
that influenced the survivability of the
crash and learn from that so that
tomorrow somebody else can benefit from
it. In the US around 50% of the vehicle
occupants who die in a car crash are not
wearing their seat
belts. So for the crash test both belted
and unbted dummies will be placed in the
cars.
For me, the most important safety
feature in our cars these days is the
seat belt. It's that seat belt that is
going to be your lifesaver.
So, what causes pileups in the real
world?
The causes of crashes are really
complex. It's not just a single factor
like speed or weather or distraction.
It's often many factors coming together.
Going too fast for the conditions is a
common feature of pileups.
So, in February of 2021 in the Fort
Worth area on Interstate 35, there was a
chain reaction crash. Uh about 133
vehicles were involved in the crash.
Multiple fatalities, half a dozen people
died, numerous people injured. One of
the likely factors in this horrific
pileup was black ice.
All of a sudden, they're in an
uncontrollable situation and once they
hit the ice, there is no
stopping. It's incidents like this that
have inspired the scenario for James'
experiment.
His plan is to replicate the conditions
of icy roads for the
crash. If we can make the road more
slippery, then you will start to see
what the effect of that will be on the
stopping distances of the cars.
To identify the perfect surface to match
the slippery conditions, James and his
team have covered a section of tarmac
with three different ice
substitutes:
gravel, oil, and water, and oil on its
own. The gravel, uh, we know that could
be a bit like marbles. Could be a little
surprise on that one. Um, oil on its
own, obviously, we're expecting that to
reduce friction, so that could skid
quite well. Um, however, I am kind of
pointing more towards the oil and water
if I'm brutally honest. But I could be
wrong. But which surface most closely
replicates the stopping distance on an
icy road? To find out, stunt driver Paul
will drive at a moderate 40 mph, then
hit the brakes when his car reaches each
surface. First up, oil. Okay, Paul. So,
the right lane, just the oil in three,
two, one, action.
[Applause]
Okay, turn your logger off. Stay there
for a second.
The next lane combines oil and water.
And because oil floats on water, the
team believes this mixture will reduce
traction even further.
Okay, Paul. So, this is oil and water in
three, two, one, action.
[Applause]
Blime me. To be honest, that's that's a
shocker. That's like a foot and a half
further than oil on its own, and that's
it.
Neither option has come anywhere close
to providing the reduced friction the
team needs. So, they are now pinning all
their hopes on the loose gravel.
Okay, Paul, stand by it. So, it's the
gravel lane. Ready? And three. Two, one,
[Music]
action. That actually is a surprise.
That That is a surprise. Okay, I'll stay
there.
Unexpectedly, loose gravel looks to be
the best substitute to replicate ice. As
soon as I touched the brakes, there was
nothing. It just went on. Was it just
constantly trying to do something? And
um whereas the other two, they felt
exactly the same and the distance was
very
The gravel reduced the grip on the road
just as freezing conditions affect
traction.
The team decides to use it in the crash
test in Scotland.
But ice on the road isn't enough.
something we'll still need to instigate
the crash.
One of the worst accident types is where
a vehicle crosses the central
reservation into oncoming traffic. And
it's this very scenario that James and
his team will aim to recreate.
We've chosen a heavy truck which will
initiate the accident.
This truck, also remotely controlled to
keep the driver safe, will cross the
center line and block the entire road,
leaving the drivers only seconds to
[Music]
react. With all the vehicles now fully
prepped, drivers chosen, and a mechanism
for the crash decided upon, the team is
ready to stage the multicar pileup.
A crew of automotive engineers and
technicians has descended onto the west
coast of
Scotland where the crash test will take
place in less than a week.
[Music]
Go very left first. That's it.
Keep it on that. That's it. Keep it on
that. Nice. Straighten up.
For the unsuspecting drivers, the true
nature of the experiment will only
become clear once the remote controlled
cars are barreling down.
This this taxiway is available to the
crash team for a
week. After that, they'll lose access.
But for now, it's been transformed into
a nearly two-mile stretch of highway
with white lines, a shoulder, and
highway grade barriers designed to
contain the
[Music]
crash. Everything must be up and running
before the drivers
arrive, including the radio
communication system.
[Music]
The radio link worked well on the test
track in Cranfield, but it could be a
very different matter in the
continuously changing atmospheric
conditions and radio signal rich
environment of a live
airport. Okay, you're in drive.
The Audi is the first car to be
tested. With the track cars on the
tarmac, all the pods are positioned in a
temporary enclosure a safe distance
away. Video screens and blackout drapes
have been fitted to enhance the
immersive experience for the
drivers. Professional stunt driver Paul
is controlling the Audi from its linked
control pod.
Yeah, we're going to go straight to 70.
I'm happy too. Going for it now.
[Music]
It's moving. given
that let me know when you're at 70,
[Music]
please 70. Excellent. Thanks, guys. 70.
Wa! Woo! We have a runner up.
The remotec controlled Audi has achieved
the target speed on its first
attempt. It's a milestone for the
team. But then
[Music]
hang on. We had a stop. Yeah, we just
had a stop. Kim, it's not responding.
No,
the Audi has automatically engaged its
brakes. A critical failure.
James needs to figure out what the issue
is and get the whole pack running before
he loses access to the test highway.
Kim, I've just e stopped it. Um, can you
just come uh up to the car and just have
a quick check, please?
They meticulously inspect data from the
test run to try and understand what
happened. For safety, the cars are
programmed to automatically apply the
brakes if the radio signal is disrupted,
suggesting that there was a loss of
communication. So, the car has to be
able to communicate continuously with
the pods or the base stations here or
they will stop. Clearly, it's a massive
issue if we're doing a drive along the
road and one of the cars stops.
Soon they zero in on the airport's
high-powered navigation beam as one of
the likely causes of the dropout. One of
the challenges of this site is that it's
an airfield and there are some active
navigational beacons relatively close to
the area. They're completely different
frequency, but they're really high
power. The radio's trying, but when they
drop out, they drop out for several
seconds. We can't work reliably. It's
not safe. This level of uncertainty
could jeopardize the whole experiment
before it's even begun.
So obviously we can't change his
location. You know, is there another
radio solution? We've tried filters, it
helps, but this radio is not enough.
This is a devastating blow to the
experiment. The loss of radio
communication isn't just a glitch, it's
a fundamental
failure. And it leaves James with no
choice. He'll have to cancel the crash
test and return to Cranfield University.
it became fairly obvious our only
solution really was to say, "Okay, we
need to postpone
this. That's a pretty disastrous thing
to have happened." But I think our
mission had always been to create the
best system we could. And if putting a
pause into it is the way to do that,
then clearly from our perspective,
that's the optimal choice.
[Music]
Right. Antenna. Okay. First one. 2.4.
Green. Right. That's good. Yep. Green.
Red.
Okay. Ready to power on.
It's taken a year and a half, but James
and his team have returned to Scotland
with a new plan and a new radio
system. One they hope will be able to
maintain connection with the cars.
That's good and sturdy.
Instead of a point-to-point system with
a single transmission path, they're
pivoting to a mesh network.
Like a cell phone network, it allows for
multiple connections between the cars
and control
pods. 14 transceivers called nodes form
the network. One on each of the eight
track cars, one back of the pod tent
feeding the control
cars, and five spread throughout the
track. Each one acts as a junction to
route data back and forth between the
pods and the cars.
This means if one path is blocked, the
signal simply reroutes and finds another
path.
Green light on it. Yeah, green is good.
The new system should handle any
interference from the airport's antenna,
identified as one of the reasons for the
radio dropout that led to the Audi's
emergency
stop. Well, that's the theory, at least.
Car is in park and we're ready. So with
the new system up and running. Okay,
James. So we've got engine started down
this end. I've got steering initialized.
10 4. We're ready to move off. 3 2 1 and
remote control car is rolling.
We need more left hand lock. More left
hand lock. Yeah, for the first time, two
cars head out onto the track together to
see if the new system can cope. We are
uh right on the edge now here. So, I'm
just going to take it gently down to the
bottom of the track. Out of hold and
brake applied.
Okay, so we're now ready to go. You want
to count us down? Yeah. 3 2 1 go.
[Music]
If either car loses communication, it
could kill the experiment once
again, but this time there'd be no
coming back for another try. Oh, nice.
That looks gorgeous. By the way,
you are picking up speed very quickly.
70 mph. 75
mph. This one's a quick one. James and
team driver Maddie have complete control
over the fastmoving
cars. Robotics and the new radio system
are working
perfectly. A big
relief. It's kind of magic. The system
that we've gone to has got this greater
robustness and so it just means we can
let the cars run. This is what happy
looks like.
[Music]
Okay, James, free to come back in your
time. Okay, Kim, we'll come back in uh
exit one over.
So, that was amazing. Two cars at 70 mph
being remotely controlled down the
motorway. Uh absolutely super. No
dropouts. Um held lane very nicely
indeed. Yeah, wonderful.
Out on the track, attention turns to the
remote control truck, the vehicle that
will instigate the
crash. So, we're using a very similar
remote control system for this. So, we
have a radio link. We have an input,
which in this case isn't a pod. It's a
second set of of driving controls that
we're using for the driver to control
this vehicle remotely. It's much better
if you can for the driver to eyeball it
directly and then he can take his queue
directly from what he sees on track.
As the unsuspecting drivers approach at
high speed, team driver Scotty will
launch the truck remotely from a
platform overseeing the
track. With steering locked, the truck
will cross the center line to block all
three lanes.
Good to go. Three, two, one.
So Scott, how did that look? It's pretty
good actually. Yeah, couldn't get any
closer to that barrier. Yeah. Perfect.
In about 6 seconds to this line. Yeah.
So lovely. This is effectively the
scenario that'll be painted in front of
the drivers and so they'll be
approaching this point. Absolutely a
worst case scenario for someone. It
really is driving away. I would not want
to be presented this.
[Music]
The next step is to pair each volunteer
driver with the car that best suits
their individual personality and driving
style. Cautious Lynn gets the
eco-friendly
Prius. Racer Luke takes the speedy Golf
GTI. New driver Caitlyn gets the
luxurious Porsche.
And laid-back surfer Tito is handed the
rugged
F-150. Now they need to learn how to
drive
[Music]
them. Come on. First up for driver
training is
Tito. Okay, that's my car right there.
Got my name on it.
Wow, this is going to be fun. Hello,
Tito. Hello, James. You ready? I'm
ready. Let's go.
The Ford F-150, popular in the US, has a
high driving position. So, the main
difference, Tito, is obviously you're
not moving, so you don't feel the
acceleration cues. You basically got to
look at the speedo, okay, to get an idea
of how quick you're going. All right.
Three, two, one. Moving on.
Rolling. Rolling in the F-150.
Look at this. This is cool. That's
pretty good, isn't it? So, you are
looking cool. Like I said, you are
looking very cool. Cool as a good
comeback. That's it. You're good. You're
good. Almost got a sensation of movement
there. Yeah, you weird. Yeah, that's it.
So, that's 50 mph already and you've
only done 500 m.
It's a strong start for driver training.
Next up is Caitlyn. She's the least
experienced of the drivers.
Oh, this is
weird. My seat's going to have to come
right forward. I can't touch the
pedals. The view is weird, isn't it?
Very strange.
It's quite quick car, isn't it? It is.
It was hard the first go. Um, but the
third time around, I got me hands around
the maneuvering. The steering was a
little bit tough, but it was fun as
well. It was exciting.
Next on the track is self-styled racer
Luke. Honestly, I'm so buzzing. I'm so
looking forward to see how fast I can
go. I want to try and get 100. In the US
in
2022, about 49% of drivers surveyed
reported exceeding the speed limit by 15
mph on a freeway in the previous
month. And 29% of fatal crashes involved
speeding drivers.
What do you think so far? Insane. See,
even doing like 50 mph, this doesn't
feel like you're doing 50 on the screen.
No, I'm getting the hang of it now.
Yeah. Yeah. That's 60. That's 70 mph
already. That's 80. You might want to
back off. 80.
Last up is Lib, the most experienced and
most cautious driver.
Okay. So, you're now in drive. All
right. Oh, and we're off. There you are.
There we go. And off you go. Bloody
Nora.
I'm going all over the bloody place. No,
no, no, no, no. You're not. No. That's
it. I don't know whether it needs
glasses. That's it. Just really small
movements. Just really, really gentle.
Tell me how fast am I going. Oh, you're
doing 30 now. Oh, is that all? Yeah. Now
you're doing 35. Oh, right. Ah, there we
go. That's more like Come on then. Way
granny rides again.
Oh dear me. Yeah, I just feel a
bit Makes you feel a bit
weird, but it's good. All good. Yeah,
really enjoyed it. Apparently, it was
going quite
fast. To create a realistic highway
pileup using remotely driven cars, it is
crucial that the volunteers are fully
immersed in the driving
experience. Experimental psychologist
Natasha Merritt is interested to see how
far this immersion goes.
[Music]
It's really interesting to see how Tito
really feels like he's driving that
truck. The way he's sitting and he's
feeling a bit higher and he's putting
his arm on the side even though he's
nowhere near that truck. You can see
that what Luke is seeing in the road.
It's quite an immersive environment. So,
it's really interesting how he's
basically using his mirrors as he's been
trained to do so in a normal driving
environment. And then go back into the
other lane. All right, let's go. That's
it. Give it an indicator. Checking the
spot. Look. Yeah, look at that. He's
totally like he's driving that that
golf. That was really really impressive
to see. Actually,
it's a good sign. The drivers are
quickly becoming immersed. The pods are
providing the realistic driving
experience that this experiment demands.
And the volunteers are still unaware of
the experiment's full objective and what
they'll really be facing the following
day.
[Music]
Hundreds of hours of planning and
preparation have all built to this
point. Now the team will discover if it
was worth
it. With fewer than 2 hours to go before
the final run and with the drivers off
site, the crash scene is prepared.
As so many pileups are caused by snow or
ice, the loose gravel selected during
the ice simulation test is laid on the
track to increase breaking
distance. The gravel will also kick up
dust and reduce
visibility. Another common feature of
pileups. Stationary vehicles are
positioned on the shoulder to create
extra hazards.
Data recorders are switched
on, dash cams are fired
up, and cameras are secured in the pods
to record the drivers as they experience
the
crash. Crash analysis expert Janet Beh
wants to see how occupants of the cars
would fare if this were a real
accident. So, she has brought along some
special
passengers. All right, guys. So, in this
vehicle, we'll have two dummies. We'll
have a dummy in the front seat,
passenger side. That dummyy's belted. In
the back, directly behind the front seat
passenger is another dummy in the rear.
He's unbted. So, my job in all of this
is to take a look at the human aspect of
the crash and what happened to the
humans because ultimately we're trying
to keep them safe during a
crash. The best scenario would have been
to put them in the driver's seat, but
today we can't do that because we have
all of this instrumentation. So, we'll
put them in the front passenger seat.
Uh, in one instance, we'll put another
one in the rear. Some will be belted,
some won't be belted. And it's all
different scenarios because that's how
the real world
is. That's great.
Unbelted. Yep. Just prop them up. Nice.
There we go.
With the cars now already, this is the
quintessential lab test with real world
combined. Frankly, we don't know what's
going to happen.
It's time to bring the drivers back to
their remote control
pods. They have no idea what is about to
happen. As far as they're concerned,
this is just the next driving test.
Hello everybody.
Welcome back. Hello. Hello. Hello. Now,
what we're going to do is three lanes
together. Okay. So, what we want to do
is drive up to 70 just as if you're in a
motorway situation. And then we're just
literally trying to simulate a motorway
drive this time. All right. Can I go in
the slow lane? Racer.
Oh, it's good to be back.
Kaitlyn
[Music]
Kyoo. Yoohoo.
All the drivers seem to have bought
James' story and are excited to drive as
a pack at high speed.
Jet, is there a peep harm? I'm afraid
there isn't any.
Four drivers from the engineering team
take the remaining pods.
Mark in the Audi and Chunwi in the Dodge
will drive at the front of the pack at
highway
speeds while Aisha in the Mercedes and
Pete in the white van bring up the rear.
Apex cameras confirm all cameras are
rolling and good to go, please. Apex
confirm green to go.
Out on the track, the truck is fired up
and team driver Scotty gets ready to
engage the vehicle from his viewing
platform. Okay, let's launch drones,
please. Drones going
[Music]
live. Drones are alive.
The team is about to find out how the
drivers react when a 36tonon truck
surprises them by careening across the
highway. All drivers, can you please
apply the brake pedal
hard? We have now given you all control.
Driving remotely, they have no idea what
lies ahead.
Now we will proceed along the motorway.
Drive up to 70 mph. Please remember,
keep in your
lane. Okay, track is live. Cars on the
move in 5 4 3 2 1. Action all vehicles.
Bloody hell, the Porsche's off. Blime
me. Everyone's steaming away. No, I
can't.
800 m.
Very good. Keep in your lane.
600 m.
400 m.
Oh. Oh, he's veering. He's
veering. There we go. There we
go. 200 m out.
What the hell's that?
Oh.
[Applause]
Oh. What? Oh my goodness.
I've just crashed. My god.
[Music]
Drivers, there has been an accident on
track. No cause for alarm. Everybody is
safe. We will now investigate. What was
that for real or was that AI? Was
Caitlyn's car gone underneath the truck?
Deep breaths, Caitlyn. Deep breaths.
The team has completed the first stage
of the experiment. A multi-car pileup
with no injuries. Every detail
painstakingly recorded from start to
finish.
JB team, please kill cars. Over. Team
leader James Brighton is now ready to
begin the next phase of the experiment.
analyzing the incredible footage, vast
quantities of data, crash scene, and
witness
accounts before the cars are sent back
to Cranfield University, the team's
base, where they'll be further analyzed
by automotive engineering
students. After years of preparation,
it's a huge relief to see all the pieces
finally come together.
This is so great. So cool.
Helping to dissect the pileup is crash
analysis expert Janet Bah.
So the golf made it through. Yeah. It it
clips the back end. Wow. Y for Janet,
this wreckage site has extraordinary
potential.
Normally she only gets to see the
aftermath of a real life crash, but this
crash is different.
Each vehicle was equipped with black
boxes that captured every detail of the
event, including speed, impact force,
and brake
pressure. Plus, with more than 90
cameras recording the entire incident,
Janet and her team will have a 360° view
of how the scene unfolded.
I'll tell you what, this is the coolest
thing
ever. So neat to see this and to have a
video of it the entire way through is
fantastic.
Although it's rare for cars to catch
fire in crashes, the fire crew makes
sure the site is
[Music]
safe. Back at the pod enclosure, the
four volunteer drivers, Tito, Caitlyn,
Luke, and Lynn, are still processing
what just happened. I saw Luke slow
down. I was like, "Oh, I I bet I got to
slow down." Oh my god, I'm so confused.
They were selected to represent
different ages and driving styles. None
of them knew they were about to be
involved in a multicar
pileup. Before taking part, they all
underwent psychological screening to
make sure they could cope with the
emotional toll of being involved in the
crash and will now be assessed again by
a psychologist.
None of these assessments are filmed in
order to protect the driver's privacy
and
well-being. The other four drivers,
Mark, Aisha, Pete, and Chunwi, are part
of James' team and knew exactly what was
coming. Hi everybody. All four of the
volunteer drivers are given the all
clear, and they all want to learn more
about the experiment and see the crash
site up close. So what we have here and
what all of these people have put
together was done on purpose. We thank
you for being a part of this. We used
you as the drivers who didn't know what
was going on. And my job in all of this
is to look at how you fared in your
vehicle. Thank god you weren't actually
in it. We have in our world, we have
crash tests where it's all very
sanitized and instrumented and we know
exactly what's going to happen for the
most part before it does. Yeah, in the
real world, of course, we have no idea.
What we did is we mixed the two worlds.
A lot of naysayers about how this is
never going to work. We did it. We did
it. Thank you. You were a huge part of
this. Oh, dude.
Not one car escaped the pileup without
damage. Some are almost unscathed.
Others are barely recognizable.
One element of it that we really didn't
want is just one enormous pileup where
there is no decision to be made by the
drivers. There's really very little to
learn. Every driver took a different
reaction to the events that were
unfolding in front of
them. In the real world, every major
pileup is carefully investigated to
determine what happened.
This staged crash is no
different. A team of crash investigators
will now analyze the incident using
standard forensic
techniques. To keep the investigation as
realistic as possible, the forensic team
has been sequestered. They did not
witness the crash and have no prior
information.
[Music]
Leading the crash forensic team is
former US state trooper Andy
Shelton. So, I'm going in blind. Um, and
I'll approach this crash as I would any
other. Uh, I'll work my way to the
center of the chaos and try and uh
travel the paths back outward until I
can figure out uh some idea of what's
going on and hopefully be able to uh
give you a cogent report at the end of
it.
Andy spent more than 20 years with the
Tennessee Highway Patrol, including time
with their critical incident response
team, where he responded to hundreds of
incidents. But he's never investigated a
pileup that's been captured in minute
detail by so many
cameras. This unique opportunity is a
chance to put traditional crash
investigation techniques to the test.
The challenge is you have the disorder
or the chaos of the crash scene, but you
have to be able to make order out of
it. The mantra in forensics is every
contact leaves a trace. Uh we're able to
use that to figure out how they came
into the crash and and ultimately what
caused
it. For the investigation, Andy is
pairing up with Marcus Row, a former UK
police forensic collision
investigator. He too saw nothing of the
accident. I always wanted to be
somebody's lackey.
Their first job is to scan and model the
entire scene in 3D using lidar. Yeah.
No, you got line of sight there.
Lidar technology allows investigations
to continue long after the wreckage has
been removed and roads reopened.
So, we're going to be able to take the
scene that we have here and drop it back
onto, for instance, Google Earth or
something like that. If this were to be
a criminal case, we can show the context
of the roadway around it. Uh things like
line of sight issues, hills, trees,
anything that may have distracted a
driver along the way. Okay, that's good.
With the entire crash scene scanned, the
forensic investigation begins.
So, you can see the pulsing from the ABS
on the on the roadway. And if you look
here, Marcus, there's some scuffing
here, and then we've got the flat tire
scuff from the left front tire.
As in the real world, the challenge for
the crash investigators is to try and
figure out how the pileup unfolded. So,
this could be undercarriage of the Audi.
giving it trailers like this.
Janet, who knows exactly how the crash
played out, will be able to judge how
accurate their conclusions are. Those
marks are to the far side tires. We're
looking for the the physical evidence
that talks to us, right? That tells us
the story that can only be made in one
way.
[Music]
As Andy and Marcus work to pick apart
the crash details, Janet can now show
the drivers how their actions in the
pods played out on the track and whether
the safety systems in their cars would
have done enough to protect
them. First up is Lynn and her
Prius. Lynn was driving at the rear of
the pack with team driver Pete following
closely behind in the white van. From
her position, she's able to spot a
problem ahead and quickly brings her
Prius to a stop.
Whoa. What's bloody happened there? Oh
god. Pete, traveling below the speed
limit at 62 mph plays the role of a
distracted driver and crashes into the
back of her
car. Oh
god. And I thought I was safe. I didn't
realize that had happened. Oh my god.
Oh. Crying out loud. I didn't think I'd
feel like this. And it's not real. But
it could have been. Oh yeah.
Despite the Prius being rearended, a
built-in safety feature called a crumple
zone took the brunt of the damage.
Crumple zones are designed to deform
under loading so that other areas of the
vehicle are better protected in a crash.
You're transferring less crash energy to
the occupants and therefore you're
reducing the injury risk.
Oh, my grandkids have survived. Oh, I
know. Let's go take a look.
Data from the onboard crash recorders
registered an impact of
22.5G when the Prius was rearended by
the van.
What's going to happen there? Oh god. To
put that in context, astronauts
experience up to 4G during takeoff. The
goal is always to keep the occupant
cabin
intact. And so, you want to dissipate
those crash energies elsewhere in the
vehicle. This helps to reduce damage to
these important areas of the vehicle
that could lead to things like fires or
leaks or other failures. This is what's
underneath. This is crumple zones are
designed to absorb forces of around 20g.
The 22.5G impact has totally crushed the
trunk space and led to a deformation of
the rear cabin. What injuries would I
have sustained? You think? So, it's hard
to say, but I think you might have some
bruising, abdominal bruising from the
seat belt.
Um, probably some neck pain at the very
least. Um, I'm a little bit more worried
about somebody in the back cuz they were
directly impacted. Ah,
the rear of the car has crumpled up to
the back seats, highlighting the intense
force the Prius
experienced. However, the deformation
inside the rear passenger cabin was
minimal, meaning severe injuries would
have been
unlikely. Still, the importance of
wearing a seat belt in the rear seat
couldn't be clearer.
Your babies weren't in there. You're
good. Hey, I know. I know there if they
were. I know. I never thought I'd feel
like this. I thought it was just a game.
Do you know what I mean? But it's like
real.
It's not. But it it Yeah. It's really
got me flaming Nora. Yeah.
The Prius is a hybrid
car. If damaged, the battery could catch
fire, posing an additional risk to
occupants and first responders.
[Music]
But in this case, the battery remained
intact. The car's crumple zones absorbed
the impact as designed, protecting the
battery. As a result, there was no risk
of a battery
fire. But can Andy and Marcus figure out
the scenario that led to the Prius
getting hit using only the evidence left
at the scene?
It's apparent that someone probably
overreacted to the crash occurring in
front of them and then possibly an
inattentive driver drove into the back
of them. So that's that's kind of a a
clear and defined crash that's separate
from everything else that we're looking
at. Tito, an experienced US driver who's
based in the UK, spots the danger early.
Remotely driving a Ford F-150 pickup,
Tito reacts quickly even though the
flying grit has reduced visibility.
What the hell's that? Oh my
goodness. What?
Tito may have benefited from the higher
driving position of the truck.
Taller vehicles can have some safety
advantages, including the fact that you
just have better visibility. You're
riding higher. you're able to see more.
They often are also larger and heavier,
and this has benefits in a crash because
of just basic
physics. When you're coming against
smaller vehicles, your vehicle is going
to have a safety advantage.
For the crash forensic team, this
analysis is straightforward.
That side of this section of the road is
higher than this side. So, these are
locked tire marks and they're tracking
directly in a straight line. the car did
not yours. So, the F-150 is coming in
straight line and it's just followed
this curved path and it's just kind of
come to a rest gently there.
[Music]
But is this how Tito remembers the
crash? Wow. Oh my gosh. Look at that.
Look at your car.
It's all right.
Did you mean to to move over, dude? I I
was just like just hold holding on
because I didn't want to hit anyone.
Okay. Wow. You could drive away, right?
Right. It looks like it. No. Tires
intact. Yeah. So, this is interesting
because it's a higher frame car. Yep.
Uh, a lot of trucks in the US. Yes. And,
um, you know, we we've got some
interesting cargo. Yeah. I'm surprised
it didn't go through. You know, you
fared really good in this. Let's take a
look inside. How about the passenger?
Passenger's. Okay. So, as a belted
occupant and the passenger side here,
the dummy was jostled around. Not a big
deal. He's belted in this impact. Pretty
good. You walk away. We're good to go.
Lynn and Tito managed to avoid hitting
the truck since they both reacted
quickly to the unfolding
situation. Natasha Merritt is an
experimental psychologist researching
car safety. For her, the footage is
revealing.
As he's coming towards the crash, he
tries to avoid the crash like in the
real world. How's that? Oh my goodness.
Instead of just breaking, he actually
tries to move out the way by steering,
which is more of an experienced driver's
behavior.
Whoa, whoa, whoa, whoa. With Lynn, she's
well behind. She sees it all happening
way ahead. She slows down even more.
She's sort of avoided the crash, but as
is also again typical in these
situations, she ends up being rearended
by another vehicle, which happens again
in the rear wells.
Being rearended is a reminder that for
drivers like Lynn, danger on the road
can be behind the driver as much as in
front. You just got to be so careful of
what's behind you and what's in front of
you and what's at the side of you. You
can't just be centered on your own
little world. You've got to be aware of
everything all around you when you're on
the motorway. Especially
having shown Tito his F-150, Janet now
wants to show him one of the other cars
on the tarmac.
You had mentioned you have a Dodge
minivan. Yes. And I'll go visit the
States. That's that's what I drive.
Let's go take a look at it.
The minivan was driven by Chunwi, one of
James' team drivers. During the crash,
it experienced multiple
collisions. And from the wreck, it's
clear a major safety device came into
play.
Airbags. The airbags used in today's
cars originated in Japan in 1964.
So a Japanese engineer called Kaborisan
came up with the idea of using a
chemical impellant to create a
controlled explosion and generate a lot
of gas in a very short
time. Later airbags were developed to
inflate in around 30 milliseconds and
were introduced into high-end vehicles
in the 1970s.
Today, they're a standard safety feature
in every new
car, but they have their limitations.
Airbags are designed to deploy once
during a crash. So, for subsequent
events, they're not going to have their
full effectiveness. Automotive
manufacturers would love to be able to
design against multiple impacts for one
vehicle. It's difficult, though, because
after a first impact, an occupant might
be inside the airbag, so to speak.
If the same airbag deploys a second time
while an occupant is already cushioned
against it, the controlled explosion
could injure rather than
protect. The occupants of this minivan
would have encountered this problem
since it experiences multiple
impacts. Its first point of contact is
with the parked black Audi A6 towing the
camper.
It's at this point the front airbags
inflate, but then it slams into the
parked blue Ford C-Max, cannoning the
C-Max into the tractor trailer's
cabin. Here, the minivan's side airbags
deploy. It then smashes into the tractor
cabin itself and finally collides
sideways with the truck's
trailer. The front and side airbags
deploy during the first two
impacts. However, since they are
designed to deflate quickly to avoid
trapping or suffocating injured
occupants, they offer less protection
during the final sideon collision.
Oh my goodness, man. Yeah, this car went
through a lot. Uh it wasn't just one
impact obviously because the entire
front is
destroyed. The side is destroyed. You
can see toward the back you've got rear
impact. We had a dummy sitting in the
front passenger seat.
So, the airbags deployed, but at the end
of the day, it didn't do a whole lot.
Okay. No. Airbag or not, with this much
intrusion, with this much damage to the
occupant
space, nobody would have survived that.
No. My goodness. Just shocking that
within m just seconds really seconds.
Solid thing turns into a crushed piece
of metal. Yeah. This is the car seat
with the seat belt that held the car
seat in. Yeah. So, if there were if
there were a child sitting back there,
they didn't make it 100%. It's a huge
impact. I will continue being a
defensive driver. Good. Major impacts
like this just any car any car will get
crushed.
But can Marcus pick apart the clues in
the crushed
metal? I think the white vehicle, the
people carrier, has probably to be
confirmed with pate transfers and other
things had an impact with the blue
towing vehicle. It's