Browse content similar to Dawn of the Driverless Car. Check below for episodes and series from the same categories and more!
Line | From | To | |
---|---|---|---|
The motorcar has shrunk the world, | 0:00:03 | 0:00:06 | |
increased personal freedom, | 0:00:06 | 0:00:08 | |
and in so many ways, expanded our horizons. | 0:00:08 | 0:00:11 | |
But there's a flip side. | 0:00:12 | 0:00:14 | |
Cars have destroyed our environment, | 0:00:14 | 0:00:16 | |
poisoned the air we breathe, | 0:00:16 | 0:00:18 | |
and killed us in far more straightforward ways. | 0:00:18 | 0:00:22 | |
There's about a million deaths every year in this world due to traffic | 0:00:23 | 0:00:27 | |
accidents and I find this just | 0:00:27 | 0:00:29 | |
utterly unacceptable in the 21st century. | 0:00:29 | 0:00:32 | |
But all that's going to change. | 0:00:32 | 0:00:35 | |
Soon, we'll be in a position to have our automotive cake and eat it. | 0:00:35 | 0:00:39 | |
Self-driving cars are going to have a huge impact on society. | 0:00:40 | 0:00:44 | |
They'll be able to navigate through complex intersections | 0:00:44 | 0:00:47 | |
with no collisions. | 0:00:47 | 0:00:48 | |
If I could just sit back and read a book, listen to music, | 0:00:48 | 0:00:51 | |
catch up on some sleep, that would be great. | 0:00:51 | 0:00:53 | |
Once your car can come from round the corner, | 0:00:53 | 0:00:56 | |
we can start opening up some of these sort of residential areas and | 0:00:56 | 0:00:59 | |
they'll feel much more sociable again. | 0:00:59 | 0:01:01 | |
This is a world where cars will drive themselves, | 0:01:01 | 0:01:05 | |
a world where we are simply passengers, | 0:01:05 | 0:01:07 | |
ferried about by wholesome, green, | 0:01:07 | 0:01:09 | |
compassionate technology which will never, ever go wrong. | 0:01:09 | 0:01:13 | |
And it's almost here. | 0:01:13 | 0:01:15 | |
I can press this button. | 0:01:15 | 0:01:17 | |
I don't even have a control that I can grab. | 0:01:17 | 0:01:20 | |
A fully autonomous vehicle in commercial operation in 2021. | 0:01:20 | 0:01:26 | |
But cars that can run errands for us by themselves could quickly clog up | 0:01:27 | 0:01:31 | |
our streets and ruin livelihoods, too. | 0:01:31 | 0:01:34 | |
From taxi drivers to truck drivers, | 0:01:36 | 0:01:39 | |
lots of people have a job as a driver. | 0:01:39 | 0:01:43 | |
A lot of effort has been put into selling us the driverless dream. | 0:01:43 | 0:01:48 | |
Now it's almost upon us, | 0:01:48 | 0:01:50 | |
could we actually be sleepwalking into a nightmare? | 0:01:50 | 0:01:53 | |
What happens when it eventually encounters a no-win scenario, | 0:01:54 | 0:01:57 | |
when it actually has to have an accident? | 0:01:57 | 0:01:59 | |
We love cars. | 0:02:13 | 0:02:14 | |
We love owning them, | 0:02:14 | 0:02:17 | |
we love driving them, | 0:02:17 | 0:02:20 | |
and learning to drive a car is a rite of passage. | 0:02:20 | 0:02:23 | |
Handbrake off. Handbrake. | 0:02:23 | 0:02:25 | |
-Oop. -Handbrake. | 0:02:25 | 0:02:27 | |
You've got to be able to drive or else you're really not a real adult. | 0:02:27 | 0:02:30 | |
But learning to mirror, signal and manoeuvre is on its way out, | 0:02:32 | 0:02:37 | |
because shadowy backroom technologists at places like Google, | 0:02:37 | 0:02:40 | |
Intel and even Facebook | 0:02:40 | 0:02:42 | |
are hell-bent on getting rid of drivers altogether. | 0:02:42 | 0:02:45 | |
We're building systems to enable cars to drive themselves. | 0:02:46 | 0:02:50 | |
Our lovely cars are on borrowed time. | 0:02:54 | 0:02:57 | |
There's even a plan drawn up by the Society of Automotive Engineers. | 0:02:57 | 0:03:01 | |
A road map, if you will. | 0:03:01 | 0:03:03 | |
It starts with level zero. | 0:03:04 | 0:03:07 | |
Level zero is where the human driver has control of everything. | 0:03:07 | 0:03:11 | |
The plan then moves through various | 0:03:12 | 0:03:14 | |
levels of automation and driver-assist | 0:03:14 | 0:03:17 | |
technologies and ends up at level five. | 0:03:17 | 0:03:20 | |
Level five is where humans are simply passengers. | 0:03:21 | 0:03:24 | |
At this point, the word "driving" | 0:03:25 | 0:03:28 | |
reverts to being something to do with livestock. | 0:03:28 | 0:03:31 | |
If this sounds a bit like a science fiction writer's pipe dream, | 0:03:33 | 0:03:36 | |
then remember, an awful lot of effort and, crucially, money | 0:03:36 | 0:03:40 | |
is currently going into it. | 0:03:40 | 0:03:42 | |
Google is investing 30 million annually to driverless, | 0:03:43 | 0:03:47 | |
and Intel recently paid 15 billion | 0:03:47 | 0:03:50 | |
for Israeli driverless tech company Mobileye. | 0:03:50 | 0:03:53 | |
From niche robot race cars... | 0:03:55 | 0:03:58 | |
Look, it's being steered by nobody! | 0:03:58 | 0:04:00 | |
..to mass manufacturers, | 0:04:01 | 0:04:04 | |
driverless is where it's at. | 0:04:04 | 0:04:06 | |
I think the promise of a driverless future where cars are available to | 0:04:07 | 0:04:12 | |
everyone and they can do all the | 0:04:12 | 0:04:14 | |
hard work and ease congestion and ease pollution is very exciting, | 0:04:14 | 0:04:19 | |
but I am a bit doubtful as to | 0:04:19 | 0:04:20 | |
whether that's how it will actually play out. | 0:04:20 | 0:04:23 | |
Autonomy is probably the biggest | 0:04:23 | 0:04:25 | |
thing that's being talked about in the automotive industry. | 0:04:25 | 0:04:28 | |
People that are less able will be able to get around because they | 0:04:28 | 0:04:31 | |
won't need to be driving. But obviously, | 0:04:31 | 0:04:33 | |
there's so much legislation to get in place to get to the point where | 0:04:33 | 0:04:36 | |
-this is actually viable. -I do think people want... | 0:04:36 | 0:04:39 | |
In principle, they want to be able to move around with less effort. | 0:04:39 | 0:04:42 | |
I think everybody wants that. | 0:04:42 | 0:04:44 | |
I'm not sure that's what Ford is talking about when it says it's | 0:04:44 | 0:04:46 | |
going to have a driverless car by 2020, or whatever it is. | 0:04:46 | 0:04:49 | |
It sounds exciting, | 0:04:50 | 0:04:52 | |
but I think it's... | 0:04:52 | 0:04:55 | |
I want to make sure it works, and that's the biggest challenge. | 0:04:55 | 0:04:58 | |
Is it going to do what it's supposed to do? | 0:04:58 | 0:05:01 | |
Will I trust it? | 0:05:01 | 0:05:03 | |
Whatever eventually emerges onto the roads of tomorrow, | 0:05:03 | 0:05:06 | |
the future looks bleak for the intimate relationship we currently | 0:05:06 | 0:05:09 | |
enjoy with our fine, four-fendered friends. | 0:05:09 | 0:05:12 | |
This is a momentous day in Harrison's life. | 0:05:20 | 0:05:24 | |
Today he is having his first driving lesson. | 0:05:24 | 0:05:27 | |
I learned to drive when I was 17. | 0:05:28 | 0:05:31 | |
I was quite nervous. | 0:05:31 | 0:05:33 | |
I loved learning to drive. I'd been | 0:05:33 | 0:05:34 | |
looking forward to it for quite a long time. | 0:05:34 | 0:05:36 | |
I was failed pretty much about | 0:05:36 | 0:05:38 | |
15 minutes in for failing to give way to someone. | 0:05:38 | 0:05:40 | |
Harrison has never attempted to drive before, | 0:05:40 | 0:05:43 | |
and now he's nailed the | 0:05:43 | 0:05:44 | |
all-important selfie with his new best friend, | 0:05:44 | 0:05:47 | |
there's a lot to think about. | 0:05:47 | 0:05:49 | |
Push that all the way down? | 0:05:49 | 0:05:50 | |
Yeah, press the brake all the way down and press the start button. | 0:05:50 | 0:05:53 | |
Look all around the car like this. | 0:05:53 | 0:05:55 | |
Just pop your signal on, press the button in, | 0:05:55 | 0:05:57 | |
pull it up and ease up off your clutch, slowly. | 0:05:57 | 0:05:59 | |
-That's it. -OK. -And we're just going to go off like that. | 0:05:59 | 0:06:02 | |
-OK. -Have you got the idea? -Yeah. | 0:06:02 | 0:06:04 | |
In a few short minutes, however, | 0:06:04 | 0:06:06 | |
Harrison is transformed from ordinary mortal to driver. | 0:06:06 | 0:06:11 | |
Exciting, isn't it? Yo, let's go! | 0:06:11 | 0:06:13 | |
-Let's go. -Don't worry, don't worry. | 0:06:13 | 0:06:16 | |
-That's fine. -OK, so he's not the finished article, | 0:06:16 | 0:06:18 | |
but in a couple of months, | 0:06:18 | 0:06:20 | |
he'll hopefully have passed his driving test and | 0:06:20 | 0:06:22 | |
will think nothing more of driving than he currently does of walking. | 0:06:22 | 0:06:26 | |
There you go. You're off. | 0:06:26 | 0:06:28 | |
You'll be driving back home in no time. | 0:06:28 | 0:06:31 | |
In simple terms, | 0:06:31 | 0:06:32 | |
Harrison has sensed the world around him and reacted appropriately. | 0:06:32 | 0:06:36 | |
It's what he's been doing all his life. | 0:06:36 | 0:06:38 | |
-Well done. -Only today, he's learning to do that via a machine. | 0:06:39 | 0:06:45 | |
Sensing our environment is something most of us take wholly for granted. | 0:06:45 | 0:06:49 | |
It's part of being human, something we're good at. | 0:06:49 | 0:06:53 | |
But little by little, cars have been getting in on the act, too. | 0:06:53 | 0:06:56 | |
Parking sensors, lane sensors, cruise control, | 0:07:03 | 0:07:08 | |
adaptive cruise control, | 0:07:08 | 0:07:10 | |
automatic headlights and windscreen wipers have all emerged under the | 0:07:10 | 0:07:14 | |
banner of driver-assist technologies. | 0:07:14 | 0:07:16 | |
But in reality, they are the first, | 0:07:18 | 0:07:21 | |
small steps on the road to full autonomy, and that's official. | 0:07:21 | 0:07:25 | |
It's the first stage on the driverless masterplan. | 0:07:25 | 0:07:29 | |
Level one. | 0:07:31 | 0:07:34 | |
At level one, the vehicle can take control of individual functions like | 0:07:34 | 0:07:38 | |
acceleration or braking. | 0:07:38 | 0:07:40 | |
Effectively, cars can actually drive | 0:07:41 | 0:07:44 | |
themselves to a small degree already. | 0:07:44 | 0:07:46 | |
Electronic stability control or ABS is one of the systems that's been | 0:07:47 | 0:07:52 | |
put in that stops me from skidding and crashing. | 0:07:52 | 0:07:55 | |
Adaptive cruise control brakes and speeds up for you, | 0:07:55 | 0:07:58 | |
depending on what the car in front | 0:07:58 | 0:08:00 | |
is doing, and that is just so great for long journeys. | 0:08:00 | 0:08:02 | |
Like, I drove to Wales on the M4 the other week and it's a | 0:08:02 | 0:08:05 | |
really long journey and it was just effortless because of that. | 0:08:05 | 0:08:09 | |
Many of these automotive technologies | 0:08:09 | 0:08:11 | |
have their origins at the racetrack. | 0:08:11 | 0:08:13 | |
And even though car racing is about as driver-focused | 0:08:17 | 0:08:20 | |
as it's possible to be... | 0:08:20 | 0:08:22 | |
..there is some common ground | 0:08:24 | 0:08:26 | |
between Formula 1 and the driverless future. | 0:08:26 | 0:08:28 | |
My name is Teena Gade, and I work here at Sahara Force India Formula 1 | 0:08:32 | 0:08:35 | |
team as a vehicle science engineer. | 0:08:35 | 0:08:37 | |
The best place normally to develop a car is actually to take it to a | 0:08:38 | 0:08:41 | |
track and test it. In reality, | 0:08:41 | 0:08:42 | |
we're restricted by the regulations so we're only allowed a fixed number | 0:08:42 | 0:08:45 | |
of test sessions every year. | 0:08:45 | 0:08:47 | |
What that means is we end up having to do quite a lot of it in the | 0:08:47 | 0:08:50 | |
virtual world. We model the car, the tyres, | 0:08:50 | 0:08:52 | |
the aerodynamics and the tracks and | 0:08:52 | 0:08:54 | |
we bring it all together in a driving | 0:08:54 | 0:08:56 | |
simulator for a driver to drive around and it tells us what the | 0:08:56 | 0:08:58 | |
performance is going to be. | 0:08:58 | 0:09:00 | |
One of the most interesting things for me about the concept of | 0:09:00 | 0:09:03 | |
driverless cars is if you take what I do on a daily basis | 0:09:03 | 0:09:05 | |
but we really don't understand, | 0:09:05 | 0:09:07 | |
the bit that is difficult is the driver element, because for all the | 0:09:07 | 0:09:10 | |
computational power we have to model, for example, | 0:09:10 | 0:09:13 | |
the aerodynamics or the tyres or the track, | 0:09:13 | 0:09:15 | |
the things that go on in the human brain are incredibly complicated and | 0:09:15 | 0:09:18 | |
processed unbelievably quickly. | 0:09:18 | 0:09:20 | |
So from an engineering perspective, | 0:09:20 | 0:09:23 | |
if we could have a completely driverless car it could perform the | 0:09:23 | 0:09:25 | |
same task absolutely consistently, the same from one lap to the next, | 0:09:25 | 0:09:29 | |
every time. We would get much cleaner data and actually, | 0:09:29 | 0:09:32 | |
we potentially stand to make better connections | 0:09:32 | 0:09:34 | |
as a result and learn more. | 0:09:34 | 0:09:36 | |
Today, Teena is off to see another race team, | 0:09:38 | 0:09:41 | |
one that have taken driverless technology a stage further. | 0:09:41 | 0:09:45 | |
I actually think I want to go left here. | 0:09:46 | 0:09:48 | |
No, I was completely wrong. I can go that way. | 0:09:48 | 0:09:51 | |
Roborace want to compete in Formula E, | 0:09:54 | 0:09:58 | |
but unlike other electric race teams, | 0:09:58 | 0:10:00 | |
they have decided to abandon drivers altogether. | 0:10:00 | 0:10:03 | |
These sleek racing robots will battle it out against each other and | 0:10:05 | 0:10:09 | |
eventually against human competition. | 0:10:09 | 0:10:11 | |
But for the time being, they're testing out the concept with this, | 0:10:12 | 0:10:16 | |
their development robot, DevBot. | 0:10:16 | 0:10:19 | |
We can see inside here, | 0:10:21 | 0:10:23 | |
it actually looks quite a lot like a conventional car cabin. | 0:10:23 | 0:10:26 | |
Yeah, we've got a rack that sits behind the driver. | 0:10:26 | 0:10:29 | |
What we've done is we've taken human capabilities and we're putting | 0:10:29 | 0:10:32 | |
them in silicon and software. | 0:10:32 | 0:10:34 | |
Here at Silverstone, DevBot knows | 0:10:36 | 0:10:39 | |
the circuit so well that it's capable of | 0:10:39 | 0:10:41 | |
giving high-speed tours of the racetrack to the humans it's aiming | 0:10:41 | 0:10:45 | |
-to defeat. -I'm not actually the best passenger at the best of times, | 0:10:45 | 0:10:49 | |
and so it's going to be quite strange. | 0:10:49 | 0:10:52 | |
Hold the handbrake. | 0:10:52 | 0:10:54 | |
-I would drive like that. -OK. | 0:10:54 | 0:10:56 | |
So this is it. | 0:10:58 | 0:11:00 | |
My life in the hands of some software. | 0:11:00 | 0:11:02 | |
The blue light's gone on at the front. | 0:11:02 | 0:11:04 | |
I think we're ready to go. | 0:11:04 | 0:11:05 | |
I am really very nervous. | 0:11:05 | 0:11:07 | |
It's terrifying, the first time you get in a car and you're not touching | 0:11:14 | 0:11:18 | |
any of the controls. | 0:11:18 | 0:11:19 | |
Once you've spent some time in it, | 0:11:39 | 0:11:40 | |
you get that feeling of total confidence, | 0:11:40 | 0:11:43 | |
and you realise that the machine is | 0:11:43 | 0:11:44 | |
far better than any human could ever be. | 0:11:44 | 0:11:46 | |
Oh, wow! | 0:11:46 | 0:11:48 | |
That is hard on the brakes. | 0:11:48 | 0:11:50 | |
Can I have another go? That's actually quite good! | 0:11:50 | 0:11:53 | |
One of the big appeals about Formula 1 and motorsport in general is the | 0:11:53 | 0:11:56 | |
personalities, the drivers themselves. | 0:11:56 | 0:11:59 | |
There is also a huge following for teams. | 0:11:59 | 0:12:02 | |
It's quite good in there, it's quite good, yeah! | 0:12:02 | 0:12:04 | |
If you look at Ferrari, they're one of the biggest in the world. | 0:12:04 | 0:12:07 | |
And I think what Roborace will allow is actually if people create that | 0:12:07 | 0:12:10 | |
following for the team rather than for the driver themselves. | 0:12:10 | 0:12:14 | |
Impressive as careering driverless | 0:12:15 | 0:12:18 | |
around the track at around 200kph is, | 0:12:18 | 0:12:21 | |
there's much more to racing than just pure speed. | 0:12:21 | 0:12:25 | |
A driverless race car will have a lot on its mind. | 0:12:25 | 0:12:28 | |
What we're really starting to look at is that judgment layer. | 0:12:30 | 0:12:33 | |
"What speed should I be entering this corner?" - | 0:12:33 | 0:12:36 | |
is the critical thing. | 0:12:36 | 0:12:37 | |
The path that I should be following. | 0:12:37 | 0:12:40 | |
And then moving up into sort of the tactical decision-making layer in | 0:12:40 | 0:12:43 | |
terms of - "Am I going to overtake or am I going to actually save some | 0:12:43 | 0:12:46 | |
"battery because I want to attack in about five laps' time?" | 0:12:46 | 0:12:50 | |
So what you're basically saying is | 0:12:50 | 0:12:51 | |
that actually the car can perform the driving function as a human can? | 0:12:51 | 0:12:55 | |
Yes, that's exactly right, yeah. | 0:12:55 | 0:12:57 | |
It is Roborace's plan to build such | 0:12:58 | 0:13:01 | |
a clever driverless car that the human | 0:13:01 | 0:13:03 | |
opposition will be ground into the dust. | 0:13:03 | 0:13:06 | |
But so far, the DevBots are only racing each other. | 0:13:06 | 0:13:09 | |
With mixed results. | 0:13:09 | 0:13:11 | |
And this is the problem. | 0:13:14 | 0:13:16 | |
Robots, even quite advanced ones | 0:13:16 | 0:13:19 | |
like the ones that run DevBot, are, well, limited. | 0:13:19 | 0:13:24 | |
Here is a representative selection. | 0:13:28 | 0:13:30 | |
Looking at these brand ambassadors, | 0:13:32 | 0:13:34 | |
the future for autonomous cars doesn't look over-rosy. | 0:13:34 | 0:13:37 | |
If they stand any chance of gaining our trust, | 0:13:38 | 0:13:42 | |
they're going to have to deliver a lot more than these. | 0:13:42 | 0:13:45 | |
And DevBot. | 0:13:46 | 0:13:48 | |
Yeah, so we're turning left. Ease up very slowly off the clutch. | 0:13:53 | 0:13:56 | |
And start to steer towards me. | 0:13:56 | 0:13:59 | |
To successfully avoid crashing his car, | 0:13:59 | 0:14:01 | |
Harrison will have to be able to perform a bewildering array of tasks | 0:14:01 | 0:14:05 | |
simultaneously in real time. | 0:14:05 | 0:14:08 | |
You've gone in the wrong gear, don't worry. | 0:14:08 | 0:14:10 | |
-Oh, sorry. -I've helped you out. Don't worry. | 0:14:10 | 0:14:12 | |
That's part of being a learner. | 0:14:12 | 0:14:14 | |
-OK. -Harrison will have to be able to recognise, categorise, | 0:14:14 | 0:14:17 | |
and accurately predict the likely | 0:14:17 | 0:14:19 | |
future actions of anything he sees while driving. | 0:14:19 | 0:14:22 | |
Check this mirror | 0:14:22 | 0:14:24 | |
and put your left signal on, which is downwards. | 0:14:24 | 0:14:27 | |
In the light of those | 0:14:27 | 0:14:28 | |
near instantaneously acquired pieces of information, | 0:14:28 | 0:14:31 | |
he must cause the car to safely accelerate, change direction, | 0:14:31 | 0:14:34 | |
slow down or stop, or all or none of the above. | 0:14:34 | 0:14:38 | |
Hang on, I'm just going to stop you there. | 0:14:38 | 0:14:40 | |
We just had a car coming rather quickly. | 0:14:40 | 0:14:42 | |
And bear in mind that this is a constantly updated stream of data to | 0:14:46 | 0:14:51 | |
be interpreted and acted upon by Harrison, | 0:14:51 | 0:14:54 | |
millisecond by millisecond, | 0:14:54 | 0:14:55 | |
for as long as he's behind the wheel. | 0:14:55 | 0:14:58 | |
We just had a narrow escape with that little caravan there. | 0:14:58 | 0:15:01 | |
That, in a nutshell, is what will be required of a self-driving car. | 0:15:02 | 0:15:07 | |
It too will need to understand and interact with its environment. | 0:15:07 | 0:15:12 | |
But let's not get ahead of ourselves. | 0:15:12 | 0:15:14 | |
-We're not going down that busy one, don't worry. -Yeah, good! | 0:15:14 | 0:15:17 | |
We're only just getting to level two, after all. | 0:15:18 | 0:15:21 | |
Level two. | 0:15:21 | 0:15:23 | |
Level two is the point on the road to autonomy at which the would-be | 0:15:23 | 0:15:27 | |
driverless car can control two things at the same time, | 0:15:27 | 0:15:31 | |
like steering AND braking. | 0:15:31 | 0:15:33 | |
It's a small but significant step. | 0:15:35 | 0:15:38 | |
But before we send all our current cars to the crusher, | 0:15:38 | 0:15:41 | |
it's worth remembering that some of them can multitask already. | 0:15:41 | 0:15:45 | |
Automatic parking I think people are very happy to use. | 0:15:46 | 0:15:50 | |
It scans for a car parking space, | 0:15:50 | 0:15:52 | |
and then you just have to do the | 0:15:52 | 0:15:54 | |
gas and brake while it does the steering for you. | 0:15:54 | 0:15:56 | |
All these tricks are all well and good, | 0:15:56 | 0:15:59 | |
and of course really very clever. | 0:15:59 | 0:16:01 | |
But to be fair, they're not exactly | 0:16:01 | 0:16:03 | |
difficult for human drivers to pull off. | 0:16:03 | 0:16:06 | |
It makes you wonder why we're | 0:16:06 | 0:16:08 | |
bothering with driverless cars at all. | 0:16:08 | 0:16:10 | |
The main attractive feature of | 0:16:10 | 0:16:12 | |
driverless cars is the promise that they will save lives. | 0:16:12 | 0:16:17 | |
I don't think humans are that great at driving. | 0:16:17 | 0:16:19 | |
You only have to look at the statistics to see that | 0:16:19 | 0:16:22 | |
there is a need for improvement there. | 0:16:22 | 0:16:25 | |
So if automating the driver can reduce those deaths, | 0:16:25 | 0:16:30 | |
then it's definitely a desirable thing. | 0:16:30 | 0:16:32 | |
There are systems that can detect whether you fall asleep or not. | 0:16:33 | 0:16:36 | |
And if you don't wake up, then the car will perform an emergency stop. | 0:16:36 | 0:16:40 | |
And some really cool stuff has been happening in collision avoidance, | 0:16:40 | 0:16:43 | |
so the car taking over when it | 0:16:43 | 0:16:45 | |
thinks you're going to have an accident | 0:16:45 | 0:16:48 | |
and you haven't reacted quickly enough. | 0:16:48 | 0:16:50 | |
The number of accidents that's saved alone over the last couple of years | 0:16:50 | 0:16:54 | |
is phenomenal. | 0:16:54 | 0:16:55 | |
The promise of a safer-than-human driverless car has been around for | 0:16:57 | 0:17:01 | |
almost as long as the car itself. | 0:17:01 | 0:17:03 | |
We trust our lives in machines all the time. | 0:17:07 | 0:17:10 | |
Every time you get on a commercial flight, | 0:17:12 | 0:17:14 | |
you're not being flown by a human any more. | 0:17:14 | 0:17:16 | |
And that's good, because you're safer this way. | 0:17:16 | 0:17:18 | |
Sebastian Thrun has spent most of | 0:17:18 | 0:17:20 | |
his professional life trying to bring | 0:17:20 | 0:17:23 | |
that airline level of safety to our cars. | 0:17:23 | 0:17:25 | |
I got involved because I had a traumatic event as a teenager. | 0:17:26 | 0:17:30 | |
My best friend died in a traffic | 0:17:30 | 0:17:31 | |
accident from one moment to the next. | 0:17:31 | 0:17:34 | |
And I found his death kind of a little bit ridiculous. | 0:17:34 | 0:17:37 | |
I think we don't talk about that much, | 0:17:37 | 0:17:38 | |
but there's about 1 million or 1.2 million | 0:17:38 | 0:17:41 | |
deaths every year in this world to traffic accidents. | 0:17:41 | 0:17:44 | |
And I find this just utterly unacceptable in the 21st century. | 0:17:44 | 0:17:47 | |
So I wanted to fix that. | 0:17:47 | 0:17:49 | |
Given that 90% of traffic accidents are due to | 0:17:49 | 0:17:52 | |
human error, Sebastian decided that the best and safest course of action | 0:17:52 | 0:17:56 | |
would be to dispense with the driver altogether. | 0:17:56 | 0:17:59 | |
In 2005, he had a breakthrough. | 0:18:03 | 0:18:05 | |
My team at Stanford built the car that won the DARPA Grand Challenge, | 0:18:07 | 0:18:11 | |
a desert race for a car that could drive itself. | 0:18:11 | 0:18:14 | |
The car crossed more than 200 kilometres of desert in a little | 0:18:16 | 0:18:20 | |
over six hours, with no human driver and no human intervention. | 0:18:20 | 0:18:24 | |
Sebastian was delighted. | 0:18:26 | 0:18:28 | |
The DARPA Grand Challenge winner and DevBot show a glimpse of the future. | 0:18:32 | 0:18:37 | |
But the trouble is that the world of driving doesn't usually exist on a | 0:18:37 | 0:18:40 | |
racetrack of known shape, size and camber | 0:18:40 | 0:18:44 | |
or in a relatively benign desert, | 0:18:44 | 0:18:46 | |
where the worst that could happen is cactus damage. | 0:18:46 | 0:18:48 | |
Real driving happens in the temporarily reversed Croydon one-way | 0:18:50 | 0:18:54 | |
system on a wet Thursday morning. | 0:18:54 | 0:18:56 | |
Formula 1 engineer Teena Gade has been impressed by the idea of robot | 0:19:00 | 0:19:05 | |
race cars. But like the rest of us, | 0:19:05 | 0:19:07 | |
she is keen to discover how | 0:19:07 | 0:19:08 | |
driverless cars might help with the real world. | 0:19:08 | 0:19:13 | |
Driving is pleasurable, say, for example, | 0:19:13 | 0:19:15 | |
you're driving on the west coast of California. | 0:19:15 | 0:19:17 | |
But if you do it for ten hours a week, | 0:19:17 | 0:19:19 | |
suddenly it's not that much fun any more. | 0:19:19 | 0:19:22 | |
The reality is that if there was a | 0:19:22 | 0:19:24 | |
train that went between where I lived and worked, | 0:19:24 | 0:19:26 | |
I'd probably take the train, | 0:19:26 | 0:19:27 | |
because it would mean I could be doing something more productive. | 0:19:27 | 0:19:30 | |
So if I had a machine that could do that for me, | 0:19:30 | 0:19:33 | |
then that would be a great thing. | 0:19:33 | 0:19:35 | |
That would buy me back ten hours a week. | 0:19:35 | 0:19:37 | |
That's a whole working day. | 0:19:37 | 0:19:38 | |
Today, Teena's off to visit a tech start-up company who want to help. | 0:19:41 | 0:19:44 | |
FIVE AI is the brainchild of Stan Boland, | 0:19:46 | 0:19:50 | |
who aims to turn this reasonably | 0:19:50 | 0:19:52 | |
priced electric car into the star of the driverless world. | 0:19:52 | 0:19:56 | |
So why would you call your company FIVE AI? | 0:19:57 | 0:19:59 | |
We're aiming for, ultimately, | 0:19:59 | 0:20:01 | |
the highest level of autonomy, which is level five, | 0:20:01 | 0:20:04 | |
a car that is utterly autonomous, | 0:20:04 | 0:20:06 | |
can drive anywhere without any kind of human intervention. | 0:20:06 | 0:20:10 | |
In fact, it hasn't even got provision for human intervention. | 0:20:10 | 0:20:13 | |
So we called the company FIVE AI. | 0:20:13 | 0:20:14 | |
Oh, wow, OK, that makes sense. | 0:20:14 | 0:20:16 | |
Urban driving creates the most challenges, actually, | 0:20:16 | 0:20:19 | |
for an autonomous car system. | 0:20:19 | 0:20:21 | |
You may have cyclists and pedestrians and cars, | 0:20:21 | 0:20:25 | |
complex buildings, | 0:20:25 | 0:20:27 | |
road markings that confuse, and people might come from any direction | 0:20:27 | 0:20:30 | |
and do almost anything in front of you, really. | 0:20:30 | 0:20:32 | |
Stan and the team are building their system from scratch, and today, | 0:20:34 | 0:20:38 | |
they're installing the sensors the | 0:20:38 | 0:20:40 | |
car will need to negotiate the world. | 0:20:40 | 0:20:43 | |
It needs to be able to see, | 0:20:43 | 0:20:45 | |
so we use cameras for that and we use light radar. | 0:20:45 | 0:20:48 | |
We also use ultrasound, | 0:20:48 | 0:20:50 | |
just like we would in our cars at | 0:20:50 | 0:20:53 | |
home to detect objects that are very, very close by. | 0:20:53 | 0:20:55 | |
But when you get to sort of solve problems like fog or | 0:20:55 | 0:20:59 | |
night-time, or rain or snow, | 0:20:59 | 0:21:02 | |
then it requires infrared cameras or high-sensitivity cameras, | 0:21:02 | 0:21:05 | |
so we should be able to sort of | 0:21:05 | 0:21:07 | |
drive along and just reconstruct the world as we go. | 0:21:07 | 0:21:10 | |
Just like humans do, really. | 0:21:10 | 0:21:11 | |
So the vision systems are actually just the tip of the iceberg. | 0:21:11 | 0:21:14 | |
What you're talking about is going on to then what's happening in the | 0:21:14 | 0:21:17 | |
-brains of this car? -That's right, | 0:21:17 | 0:21:19 | |
it does require a huge amount of cognitive capability in the car to | 0:21:19 | 0:21:22 | |
be able to sort of sense the world and then take decisions about how we | 0:21:22 | 0:21:26 | |
control it. So that is a huge problem, | 0:21:26 | 0:21:29 | |
but one that we think we can actually solve. | 0:21:29 | 0:21:33 | |
To do that, Stan and the team will need to get their system through all | 0:21:33 | 0:21:36 | |
the levels of autonomy, including level three. | 0:21:36 | 0:21:40 | |
Level three. | 0:21:43 | 0:21:45 | |
At level three, the masterplan | 0:21:46 | 0:21:48 | |
states that safety-critical functions can | 0:21:48 | 0:21:51 | |
be completely assigned to the vehicle under certain driving or | 0:21:51 | 0:21:54 | |
environmental conditions... | 0:21:54 | 0:21:55 | |
..but that a supervising human driver | 0:21:56 | 0:21:59 | |
must be present to take over in emergencies. | 0:21:59 | 0:22:02 | |
What that means in reality is that if the driving and environmental | 0:22:03 | 0:22:07 | |
conditions allow, you can shout at | 0:22:07 | 0:22:09 | |
the children face-to-face while the car drives itself. | 0:22:09 | 0:22:13 | |
Now, this is much more like it, and what's more, | 0:22:13 | 0:22:17 | |
it's almost available now. | 0:22:17 | 0:22:19 | |
And this is what it might look like in reality. | 0:22:20 | 0:22:23 | |
Or at least, what it looks like in the mind of whichever creative Volvo | 0:22:23 | 0:22:27 | |
hired to create this. | 0:22:27 | 0:22:29 | |
Once the two green bars in the centre meet, | 0:22:29 | 0:22:32 | |
the paddle lights shift to green and the autopilot confirms that the | 0:22:32 | 0:22:36 | |
driving and the supervision is delegated to the car. | 0:22:36 | 0:22:39 | |
And this is what this slightly | 0:22:42 | 0:22:43 | |
nervous-looking man from Nissan unleashed | 0:22:43 | 0:22:46 | |
on some carefully selected parts of east London earlier this year. | 0:22:46 | 0:22:50 | |
This is a real car and these are real streets. | 0:22:51 | 0:22:54 | |
And those are real hands, | 0:22:54 | 0:22:56 | |
poised to grab the autonomous steering wheel at a moment's notice. | 0:22:56 | 0:23:00 | |
Being in an autonomous car is | 0:23:01 | 0:23:03 | |
obviously a weird feeling because you're doing nothing. | 0:23:03 | 0:23:07 | |
It's quite unnerving the first time you do it, | 0:23:07 | 0:23:09 | |
certainly handing over control on a motorway, for example, | 0:23:09 | 0:23:12 | |
when it's obviously high speed. | 0:23:12 | 0:23:14 | |
But what's surprising is how quickly you become accustomed to the fact | 0:23:14 | 0:23:17 | |
that the car is driving itself. | 0:23:17 | 0:23:20 | |
Driverless cars are kept on the | 0:23:22 | 0:23:24 | |
straight and narrow using three things. | 0:23:24 | 0:23:26 | |
First, satellite navigation knows, more or less, where the road is. | 0:23:26 | 0:23:31 | |
Lidar, a spinning laser, | 0:23:31 | 0:23:33 | |
builds up an overview picture of the environment by recording its own | 0:23:33 | 0:23:37 | |
reflection, and radar does a similar thing for the short range. | 0:23:37 | 0:23:40 | |
It's this combination that Sebastian Thrun and his team used | 0:23:42 | 0:23:45 | |
to autonomously navigate through the Mojave Desert | 0:23:45 | 0:23:48 | |
and win the DARPA Challenge. | 0:23:48 | 0:23:50 | |
But for autonomously driving you home from a drink-and-drugs-fuelled | 0:23:52 | 0:23:55 | |
weekend with your crazy friends on the coast, it just can't cut it. | 0:23:55 | 0:23:59 | |
You still need a human, and that's a problem... | 0:24:00 | 0:24:02 | |
..one that's being addressed at Stanford University in California. | 0:24:04 | 0:24:07 | |
I'm going to get over this hill and then I'll talk to you. | 0:24:10 | 0:24:15 | |
I'm Wendy Ju and I work on research particularly around how people are | 0:24:15 | 0:24:19 | |
going to interact with automation. | 0:24:19 | 0:24:21 | |
So we're basically borrowing a little piece of the possible future | 0:24:21 | 0:24:25 | |
and then running experiments to see how people react there. | 0:24:25 | 0:24:28 | |
So right now, we're in a driving simulator. | 0:24:28 | 0:24:31 | |
In this particular set-up, we are trusting | 0:24:31 | 0:24:33 | |
this robot steering wheel. | 0:24:33 | 0:24:35 | |
I can press this button. | 0:24:35 | 0:24:37 | |
Now the car is going to switch to autonomy, | 0:24:38 | 0:24:40 | |
but the steering wheel handle actually goes back. | 0:24:40 | 0:24:43 | |
So if I'm a little distracted, daydreaming and I look down, | 0:24:43 | 0:24:46 | |
I know who's in charge, | 0:24:46 | 0:24:47 | |
not only because there's this little icon on the dashboard, | 0:24:47 | 0:24:50 | |
but I don't even have a control I can grab. | 0:24:50 | 0:24:52 | |
The problem for the driver, though, | 0:24:53 | 0:24:55 | |
is paying attention when the machine is in control. | 0:24:55 | 0:24:59 | |
Human minds tend to wander, get distracted, or simply shut down. | 0:24:59 | 0:25:03 | |
The idea of autonomous cars is really, really exciting. | 0:25:04 | 0:25:08 | |
So it's ironic that the reality of being in an autonomous car is that | 0:25:08 | 0:25:13 | |
being in an autonomous car is really, really boring. | 0:25:13 | 0:25:16 | |
It goes against intuition, | 0:25:16 | 0:25:18 | |
but what our research points to is how to keep people engaged and alert | 0:25:18 | 0:25:22 | |
and aware of what's going on on the road. | 0:25:22 | 0:25:25 | |
Because when we run an experiment, | 0:25:25 | 0:25:27 | |
people who are told just to | 0:25:27 | 0:25:29 | |
supervise the car are falling asleep. | 0:25:29 | 0:25:31 | |
Just the task of supervising the car | 0:25:31 | 0:25:34 | |
is not engaging enough to keep people awake. | 0:25:34 | 0:25:37 | |
And this is the Achilles heel of level three autonomy. | 0:25:37 | 0:25:40 | |
There needs to be a human ready to avert disaster at all times. | 0:25:40 | 0:25:45 | |
So we have this camera here pointed at my face to tell if they're awake | 0:25:45 | 0:25:50 | |
or asleep, if they're alert or not. | 0:25:50 | 0:25:53 | |
I think that it's going to become important for the car to do active | 0:25:53 | 0:25:56 | |
intervention. Maybe talking to people more, | 0:25:56 | 0:25:58 | |
entertain them more if they look drowsy or asleep. | 0:25:58 | 0:26:01 | |
Ask them questions, you know, | 0:26:01 | 0:26:02 | |
"Do you think we should go this way or that way?" | 0:26:02 | 0:26:05 | |
You know, "What do you think of this thing over there?" | 0:26:05 | 0:26:07 | |
If only the humans inside had something to do. | 0:26:07 | 0:26:10 | |
Like steer. | 0:26:10 | 0:26:11 | |
Just a thought. | 0:26:13 | 0:26:14 | |
I think people want fully driverless cars that can do absolutely | 0:26:16 | 0:26:20 | |
everything without human input. | 0:26:20 | 0:26:23 | |
I'm less convinced that people will | 0:26:24 | 0:26:26 | |
be happy with a supposedly driverless | 0:26:26 | 0:26:28 | |
car that actually needs them to be | 0:26:28 | 0:26:30 | |
paying attention the whole time in case something goes wrong. | 0:26:30 | 0:26:33 | |
What happens when we get used to not driving for | 0:26:33 | 0:26:36 | |
long periods of time and then are asked to act quickly if something | 0:26:36 | 0:26:40 | |
happens and take control, | 0:26:40 | 0:26:42 | |
and our reactions are not fast enough to actually do that? | 0:26:42 | 0:26:45 | |
Will they be completely distracted reading a magazine and so not | 0:26:45 | 0:26:49 | |
realise they have to take control? | 0:26:49 | 0:26:51 | |
That's one of the biggest challenges. | 0:26:51 | 0:26:53 | |
Level three automation, then, is slightly pointless, | 0:26:54 | 0:26:57 | |
a kind of cleverer cruise control | 0:26:57 | 0:27:00 | |
with a bit of steering sometimes thrown in. | 0:27:00 | 0:27:02 | |
What we really need is something | 0:27:02 | 0:27:04 | |
that can take on everything a human driver can. | 0:27:04 | 0:27:07 | |
Something clever. Intelligent, even. | 0:27:07 | 0:27:11 | |
Luckily, help is at hand on the | 0:27:14 | 0:27:16 | |
other side of the San Francisco Bay Area. | 0:27:16 | 0:27:18 | |
The University of California at Berkeley is Stanford's groovier, | 0:27:20 | 0:27:24 | |
longer-haired, more radical cousin. | 0:27:24 | 0:27:26 | |
Scientists here have come up with this. | 0:27:29 | 0:27:32 | |
It may not look radical, | 0:27:32 | 0:27:34 | |
but what it can do may just deliver the driverless revolution we've all | 0:27:34 | 0:27:37 | |
been promised. | 0:27:37 | 0:27:38 | |
Brett is the Berkeley Robot for the Elimination of Tedious Tasks. | 0:27:40 | 0:27:44 | |
If you look here, we got Brett's toys. | 0:27:45 | 0:27:47 | |
Brett has learned to stack Lego blocks, | 0:27:47 | 0:27:51 | |
has learned to put caps onto bottles, | 0:27:51 | 0:27:55 | |
assembling pieces of this airplane. | 0:27:55 | 0:27:57 | |
Whilst this might seem slightly underwhelming, | 0:27:57 | 0:28:00 | |
the key here is the idea of learning. | 0:28:00 | 0:28:04 | |
Nobody has written Brett a bottle-top program or a block-stacking | 0:28:04 | 0:28:08 | |
algorithm. The robot has worked it out for himself, | 0:28:08 | 0:28:13 | |
something that until recently was the exclusive domain | 0:28:13 | 0:28:16 | |
of animals with sizeable brains. | 0:28:16 | 0:28:18 | |
I am Pieter Abbeel. | 0:28:19 | 0:28:21 | |
I'm a professor at UC Berkeley and I work on artificial intelligence for | 0:28:21 | 0:28:24 | |
robotics. Basically, | 0:28:24 | 0:28:26 | |
robots are mechanically very capable and can do a lot of things. | 0:28:26 | 0:28:29 | |
But in practice, they do very little for us. | 0:28:29 | 0:28:31 | |
What's holding them back is their lack of intelligence. | 0:28:31 | 0:28:34 | |
Or at least it was. | 0:28:37 | 0:28:39 | |
Artificial intelligence, also known as machine learning, | 0:28:39 | 0:28:43 | |
is allowing robots to acquire new skills for themselves. | 0:28:43 | 0:28:47 | |
Just like we do. | 0:28:47 | 0:28:49 | |
So imitation learning is the process of having a robot learn from | 0:28:49 | 0:28:52 | |
watching something done for it. | 0:28:52 | 0:28:54 | |
Then in the future, faced with a different situation, | 0:28:54 | 0:28:57 | |
as to tie a knot, we're going to see the robot | 0:28:57 | 0:29:00 | |
understand what to do in the new situation. | 0:29:00 | 0:29:03 | |
For this challenge, | 0:29:03 | 0:29:04 | |
the rope is relayed in roughly the same place and Brett is asked to tie | 0:29:04 | 0:29:08 | |
the knot again. | 0:29:08 | 0:29:10 | |
Right, let's see what Brett can do now. | 0:29:10 | 0:29:12 | |
Brett can't simply repeat the exact motions he's been shown because the | 0:29:14 | 0:29:17 | |
rope isn't in exactly the same place. | 0:29:17 | 0:29:19 | |
He'll have to adapt the principles | 0:29:21 | 0:29:22 | |
of the challenge to the reality of the rope's new position. | 0:29:22 | 0:29:26 | |
It's not just repeating the motions we gave. | 0:29:27 | 0:29:30 | |
It's looking at how the new | 0:29:30 | 0:29:31 | |
situation relates to the old situation and | 0:29:31 | 0:29:34 | |
then morphing the motions, | 0:29:34 | 0:29:36 | |
make them the right motions for the new situation. | 0:29:36 | 0:29:38 | |
Yeah, there, Brett did it. | 0:29:38 | 0:29:40 | |
Beautiful knot. | 0:29:40 | 0:29:42 | |
Beautiful indeed, but Brett's skills don't end at imitation. | 0:29:42 | 0:29:46 | |
He can work out how to perform a task from scratch, like a toddler. | 0:29:48 | 0:29:52 | |
A rather unattractive toddler with metal arms and an electronic brain. | 0:29:53 | 0:29:58 | |
But if he can learn, there's no reason why one day Brett, | 0:29:58 | 0:30:02 | |
or a robot like him, couldn't learn to drive. | 0:30:02 | 0:30:04 | |
But we're getting ahead of ourselves. | 0:30:06 | 0:30:09 | |
We'll specify the objective rather | 0:30:09 | 0:30:11 | |
than the strategy to achieve something. | 0:30:11 | 0:30:13 | |
So you don't have to demonstrate | 0:30:13 | 0:30:14 | |
everything multiple times for the robot to understand what to do. | 0:30:14 | 0:30:17 | |
You just give the robot an objective. | 0:30:17 | 0:30:19 | |
Then the robot just can go at it, | 0:30:19 | 0:30:22 | |
pretty much like a kid playing, | 0:30:22 | 0:30:23 | |
trying, trying, trying and, over time, | 0:30:23 | 0:30:26 | |
getting better and better. | 0:30:26 | 0:30:27 | |
This challenge requires Brett to get | 0:30:29 | 0:30:31 | |
the red cube into the box through one of the holes. | 0:30:31 | 0:30:35 | |
But he's not shown how. | 0:30:35 | 0:30:36 | |
He has no model of how his own arm works, | 0:30:38 | 0:30:40 | |
so initially the best he can do is random motion. | 0:30:40 | 0:30:43 | |
So right now... | 0:30:43 | 0:30:44 | |
As it refines that model and gets more data from its own attempts it | 0:30:46 | 0:30:49 | |
gets better and better at finding a | 0:30:49 | 0:30:50 | |
solution to get the block into the matching opening. | 0:30:50 | 0:30:53 | |
And actually, very quickly. | 0:30:53 | 0:30:55 | |
This took maybe one minute of learning | 0:30:55 | 0:30:56 | |
and it invented how to do that from scratch. | 0:30:56 | 0:30:59 | |
With imitation learning, | 0:30:59 | 0:31:01 | |
Brett simply adapts an action that | 0:31:01 | 0:31:03 | |
he's been shown, but machine learning, | 0:31:03 | 0:31:06 | |
artificial intelligence, allows him to solve a problem for himself. | 0:31:06 | 0:31:10 | |
The implications for his descendants and ours are far-reaching. | 0:31:12 | 0:31:16 | |
Maybe you want a robot to jump higher than a human has ever jumped. | 0:31:18 | 0:31:21 | |
Or you want to do something more precise than a human has ever done. | 0:31:21 | 0:31:24 | |
You can give it the objective and then it'll, on its own, try. | 0:31:24 | 0:31:28 | |
Not succeed initially, fail most of the time, | 0:31:28 | 0:31:30 | |
but over time get better and better and maybe exceed human performance. | 0:31:30 | 0:31:34 | |
When it comes to driverless cars, | 0:31:36 | 0:31:38 | |
exceeding human performance is essential. | 0:31:38 | 0:31:41 | |
After all, not many of us would want to be driven around by Brett. | 0:31:43 | 0:31:46 | |
Artificial intelligence is the | 0:31:52 | 0:31:54 | |
latest technological new kid on the block | 0:31:54 | 0:31:57 | |
and areas that seemingly have | 0:31:57 | 0:31:58 | |
nothing to do with driving are busy making | 0:31:58 | 0:32:02 | |
friends with it, unwittingly | 0:32:02 | 0:32:04 | |
preparing AI for its upcoming automotive challenge. | 0:32:04 | 0:32:07 | |
Ruby, for example, has memorised more than 100 billion web pages. | 0:32:09 | 0:32:13 | |
No human can do this. In fact, | 0:32:13 | 0:32:15 | |
it can find the right web page as | 0:32:15 | 0:32:16 | |
you're still typing your search term. | 0:32:16 | 0:32:19 | |
That is completely impossible for the human brain. | 0:32:19 | 0:32:22 | |
I mean, before machine learning we | 0:32:22 | 0:32:23 | |
used to program computers line by line. | 0:32:23 | 0:32:26 | |
But now, we can teach computers and let them learn on their own, | 0:32:26 | 0:32:30 | |
the same way people learn. | 0:32:30 | 0:32:31 | |
At the social media giant Facebook, | 0:32:33 | 0:32:36 | |
artificial intelligence underpins the whole operation. | 0:32:36 | 0:32:39 | |
Facebook today could not exist without AI. It's as simple as that. | 0:32:41 | 0:32:46 | |
Over a billion people use Facebook | 0:32:46 | 0:32:48 | |
every day and they will load their news | 0:32:48 | 0:32:50 | |
feed a few dozen times every single day. | 0:32:50 | 0:32:52 | |
And, you know, if you imagine how | 0:32:52 | 0:32:54 | |
many people it would take if you were to | 0:32:54 | 0:32:56 | |
line out all the pieces of content that are available to you every | 0:32:56 | 0:32:59 | |
day, for how to sort out how | 0:32:59 | 0:33:00 | |
relevant is this story going to be to | 0:33:00 | 0:33:03 | |
this person, you multiply that by a billion people, | 0:33:03 | 0:33:05 | |
you see that for a human, | 0:33:05 | 0:33:07 | |
this would be a task that is absolutely impossible to do. | 0:33:07 | 0:33:10 | |
Software engineers are trying to achieve artificial intelligence by | 0:33:11 | 0:33:14 | |
programming computers to process information like the human brain. | 0:33:14 | 0:33:18 | |
They have come up with artificial neural networks, | 0:33:20 | 0:33:23 | |
systems that allow what researchers call deep learning. | 0:33:23 | 0:33:26 | |
My name is Yann Lecun, I run and I | 0:33:28 | 0:33:30 | |
build a big research group at Facebook, | 0:33:30 | 0:33:32 | |
discovering new things that nobody has thought about before. | 0:33:32 | 0:33:35 | |
So when you talk to your phone and it recognises your voice, | 0:33:35 | 0:33:38 | |
it uses deep learning. When you | 0:33:38 | 0:33:39 | |
upload your photos on Facebook and Google | 0:33:39 | 0:33:41 | |
and you can index them and search them, | 0:33:41 | 0:33:43 | |
they've been recognised by deep learning. | 0:33:43 | 0:33:46 | |
Deep learning mimics the way we learn ourselves. | 0:33:48 | 0:33:51 | |
In our brains, this is done by | 0:33:53 | 0:33:55 | |
strengthening and weakening neural connections. | 0:33:55 | 0:33:58 | |
Yann has designed an artificial | 0:33:58 | 0:34:00 | |
system that does a similar thing in the virtual world. | 0:34:00 | 0:34:03 | |
When you get an image in your eyes, on your retina, | 0:34:08 | 0:34:10 | |
it goes to the back of the brain and then it's kind of processed by | 0:34:10 | 0:34:12 | |
multiple layers of neurons. | 0:34:12 | 0:34:14 | |
And that process takes about 100 milliseconds and it goes through | 0:34:14 | 0:34:17 | |
only a few layers of neurons, maybe between ten and 20. | 0:34:17 | 0:34:19 | |
So these neural nets that we're building, | 0:34:19 | 0:34:21 | |
they're called convolutional networks. | 0:34:21 | 0:34:24 | |
It's a little bit my invention and | 0:34:24 | 0:34:25 | |
they're organised in a very similar way. | 0:34:25 | 0:34:28 | |
The reason that people are excited about Yann's convolutional, | 0:34:29 | 0:34:32 | |
artificial neural networks is their | 0:34:32 | 0:34:34 | |
astonishing abilities in object recognition. | 0:34:34 | 0:34:37 | |
They can be taught, for example, | 0:34:39 | 0:34:40 | |
to look at the digital data from a photograph | 0:34:40 | 0:34:43 | |
and let you know whether or not the picture contains a dog. | 0:34:43 | 0:34:46 | |
This is done by showing the system lots of pictures of dogs, | 0:34:49 | 0:34:52 | |
while telling the artificial neural network that these are dogs. | 0:34:52 | 0:34:55 | |
By looking for common features in the pictures, | 0:34:57 | 0:35:00 | |
the system will develop its own definition of what constitutes dog, | 0:35:00 | 0:35:04 | |
so that eventually, | 0:35:04 | 0:35:05 | |
it will be able to spot a dog or the | 0:35:05 | 0:35:07 | |
fact that a picture doesn't contain | 0:35:07 | 0:35:09 | |
a dog, all by itself. | 0:35:09 | 0:35:11 | |
Of course, it's not just dogs. It could be anything - cars, people, | 0:35:12 | 0:35:17 | |
penguins, oncoming traffic. | 0:35:17 | 0:35:19 | |
It's actually better at doing this | 0:35:21 | 0:35:24 | |
than a human engineer would be at designing a system. | 0:35:24 | 0:35:27 | |
That's the surprising thing, it's | 0:35:27 | 0:35:28 | |
very humbling for an engineer to think that's the case, but it is. | 0:35:28 | 0:35:31 | |
And they work really well for image | 0:35:31 | 0:35:32 | |
recognition, for video interpretation. | 0:35:32 | 0:35:35 | |
So all the self-driving cars that | 0:35:35 | 0:35:37 | |
you see around that use a camera input, | 0:35:37 | 0:35:39 | |
they all use convolutional nets. | 0:35:39 | 0:35:40 | |
It's these systems' ability to | 0:35:40 | 0:35:42 | |
differentiate at a better than human level | 0:35:42 | 0:35:44 | |
which is at the heart of artificial intelligence. | 0:35:44 | 0:35:46 | |
It's what powers Brett, | 0:35:48 | 0:35:50 | |
the knot-tying robot, and it's going to be critical to the difference | 0:35:50 | 0:35:53 | |
between level three driverless cars that almost work | 0:35:53 | 0:35:56 | |
and level four driverless cars that really do. | 0:35:56 | 0:35:59 | |
But it's not the only thing. | 0:36:01 | 0:36:02 | |
The other thing is this. | 0:36:03 | 0:36:05 | |
Video games are massively demanding on computer power, | 0:36:06 | 0:36:10 | |
so much so that a whole new way of | 0:36:10 | 0:36:12 | |
processing is needed to handle all the data they use. | 0:36:12 | 0:36:15 | |
What evolved was this - the graphics processing unit or GPU. | 0:36:17 | 0:36:22 | |
And because of their data handling capabilities, | 0:36:24 | 0:36:27 | |
GPUs have been gleefully adopted by | 0:36:27 | 0:36:29 | |
the would-be makers of driverless cars. | 0:36:29 | 0:36:32 | |
My name's Danny Shapiro. | 0:36:34 | 0:36:36 | |
I'm the Senior Director of Automotive at Nvidia | 0:36:36 | 0:36:38 | |
and we're building systems | 0:36:38 | 0:36:40 | |
to enable cars to drive themselves. | 0:36:40 | 0:36:42 | |
This is the brain of a self-driving car. | 0:36:44 | 0:36:47 | |
We plug in cameras, radar, | 0:36:47 | 0:36:49 | |
lidar. All the sensors of the car feed into this device. | 0:36:49 | 0:36:54 | |
A CPU, which is the central processing unit, | 0:36:55 | 0:36:57 | |
you've probably heard, has dual core or quad core, | 0:36:57 | 0:37:02 | |
meaning there's two lanes or four lanes where information flows. | 0:37:02 | 0:37:07 | |
The GPU can have thousands of cores or lanes. | 0:37:07 | 0:37:10 | |
Imagine a highway with 1,000 lanes, | 0:37:10 | 0:37:13 | |
how much traffic could you push through that processor? | 0:37:13 | 0:37:15 | |
Just like the brains of human drivers, | 0:37:20 | 0:37:23 | |
the systems that control driverless cars will be | 0:37:23 | 0:37:27 | |
voracious consumers of data. | 0:37:27 | 0:37:29 | |
They will be fed with a constant stream of digits from lidar, radar, | 0:37:29 | 0:37:35 | |
infrared sensors and multiple video cameras, | 0:37:35 | 0:37:37 | |
all of which will need to be seamlessly interpreted, | 0:37:37 | 0:37:40 | |
coordinated and fed back in the form of different data to the car's | 0:37:40 | 0:37:44 | |
driving controls in real time. | 0:37:44 | 0:37:47 | |
The driverless car has to use all this to sense and interpret the real | 0:37:47 | 0:37:51 | |
world with 100% accuracy. | 0:37:51 | 0:37:54 | |
A GPU is able to process and reconstruct, essentially, | 0:37:57 | 0:38:00 | |
a three-dimensional model of everything going on around the car. | 0:38:00 | 0:38:03 | |
All that data, then, is analysed. | 0:38:03 | 0:38:06 | |
It doesn't just sense there's an object, | 0:38:06 | 0:38:08 | |
but we know exactly what that object is. | 0:38:08 | 0:38:10 | |
It could be a pedestrian on a cellphone, | 0:38:10 | 0:38:13 | |
it could be a motorcycle, it could be an ambulance. | 0:38:13 | 0:38:16 | |
British start-up FIVE AI are training | 0:38:16 | 0:38:18 | |
their car to recognise things too. | 0:38:18 | 0:38:20 | |
Formula 1 engineer Teena Gade | 0:38:21 | 0:38:23 | |
is being shown the world through the eyes of the driverless car. | 0:38:23 | 0:38:26 | |
So talk me through what we've got on the screen here. | 0:38:28 | 0:38:31 | |
The computer vision is making the machine see. | 0:38:31 | 0:38:33 | |
So here we have the live stream from the camera coming in and these are | 0:38:33 | 0:38:38 | |
representations that have been processed from that. | 0:38:38 | 0:38:41 | |
So in the first one, what we're seeing here is in real time, | 0:38:41 | 0:38:44 | |
the actual detection of buses, cars, | 0:38:44 | 0:38:46 | |
pedestrians that have been inferred by our algorithm. | 0:38:46 | 0:38:50 | |
The next one is what we'd call a segmentation, | 0:38:50 | 0:38:53 | |
which is a breaking-up of the image into something like here's where the road | 0:38:53 | 0:38:56 | |
is, here's a wall, here's a building, | 0:38:56 | 0:39:00 | |
so that the car has a very coarse awareness of what's around it. | 0:39:00 | 0:39:04 | |
And the same sorts of machine learning techniques, | 0:39:04 | 0:39:06 | |
neural networks is what will also | 0:39:06 | 0:39:09 | |
help you solve looking at intentions of | 0:39:09 | 0:39:11 | |
other road users. You can imagine, just as a human goes out, | 0:39:11 | 0:39:14 | |
they learn how other road users use by observation. | 0:39:14 | 0:39:17 | |
So you'd feed it millions and | 0:39:17 | 0:39:19 | |
millions of days of video in all different situations. | 0:39:19 | 0:39:23 | |
And the machine itself would learn how to understand their | 0:39:23 | 0:39:27 | |
movements, maybe picking up cues that to you and I, | 0:39:27 | 0:39:30 | |
we would never have even thought of, | 0:39:30 | 0:39:31 | |
because it would have so much data at its disposal. | 0:39:31 | 0:39:36 | |
But there is a fascinating short cut to this, | 0:39:36 | 0:39:39 | |
one that Teena exploits in her work. | 0:39:39 | 0:39:41 | |
You might want to turn round. | 0:39:41 | 0:39:43 | |
I don't know what's up here. | 0:39:43 | 0:39:45 | |
And one that FIVE AI are making full use of. | 0:39:45 | 0:39:48 | |
This, actually, looks quite familiar to me. | 0:39:48 | 0:39:50 | |
This is a simulated environment, | 0:39:50 | 0:39:52 | |
and that's how we test our cars on the track. | 0:39:52 | 0:39:54 | |
Yes. So one of the exciting things about the development in gaming | 0:39:54 | 0:39:58 | |
engines and simulations of reality is they're getting so good... | 0:39:58 | 0:40:01 | |
I mean, if you look at, say, Grand Theft Auto and games like this, | 0:40:01 | 0:40:05 | |
it's now become possible to | 0:40:05 | 0:40:07 | |
actually do testing in these virtual worlds | 0:40:07 | 0:40:10 | |
which is almost as good as testing in reality. | 0:40:10 | 0:40:12 | |
And you could get algorithms up to a | 0:40:12 | 0:40:14 | |
sufficient case on very rare test cases | 0:40:14 | 0:40:17 | |
that you just wouldn't have access to. | 0:40:17 | 0:40:19 | |
You know, creating accidents, things like this, | 0:40:19 | 0:40:21 | |
which you wouldn't want to do in the real world, and actually, you know, | 0:40:21 | 0:40:25 | |
it takes a lot of money to run a car out in the real world for, like, | 0:40:25 | 0:40:28 | |
90 million miles. | 0:40:28 | 0:40:30 | |
If you have enough computers, you can do it in the virtual world | 0:40:30 | 0:40:33 | |
in, you know, hours. | 0:40:33 | 0:40:35 | |
It's the modern blurring of the real | 0:40:36 | 0:40:38 | |
with the digitally virtual that means | 0:40:38 | 0:40:41 | |
that now, right now, in the early 21st century, | 0:40:41 | 0:40:45 | |
it might be that the driverless car's time has come. | 0:40:45 | 0:40:48 | |
Most of the ideas about AI have been around for quite a while. | 0:40:51 | 0:40:54 | |
What we now have is the technology. | 0:40:54 | 0:40:56 | |
So we have the computational power | 0:40:56 | 0:40:57 | |
to run them as well as the ability to | 0:40:57 | 0:40:59 | |
gather the data in order to learn | 0:40:59 | 0:41:01 | |
what we need to learn and to train these systems. | 0:41:01 | 0:41:03 | |
So we're at this point now where driverless cars and driving | 0:41:03 | 0:41:06 | |
simulators can do both of these things, | 0:41:06 | 0:41:08 | |
because the technology enables us to at the moment. | 0:41:08 | 0:41:11 | |
And that's certainly what the car industry believes. | 0:41:11 | 0:41:14 | |
Welcome to the world of level four. | 0:41:14 | 0:41:16 | |
Level four. | 0:41:19 | 0:41:21 | |
According to the plan, | 0:41:21 | 0:41:23 | |
a level four car will perform all | 0:41:23 | 0:41:25 | |
driving functions and monitor roadway | 0:41:25 | 0:41:28 | |
conditions for an entire trip within | 0:41:28 | 0:41:30 | |
the operational designed domain of the vehicle. | 0:41:30 | 0:41:33 | |
So this is it. | 0:41:35 | 0:41:37 | |
This is, to all intents and purposes, the future... | 0:41:37 | 0:41:40 | |
..driverless cars that will ferry you around with no need for you to worry | 0:41:41 | 0:41:45 | |
yourself with troublesome bits of kit like steering wheels, or brakes, | 0:41:45 | 0:41:49 | |
or gears, or anything, really. | 0:41:49 | 0:41:53 | |
So just imagine, you go to your smartphone, you say, | 0:41:53 | 0:41:56 | |
"I need a ride to the pizza place," and you punch in the app, | 0:41:56 | 0:42:00 | |
and the car pops up in front of your house, empty. | 0:42:00 | 0:42:03 | |
You hop inside, it drops you at the restaurant, you have dinner, | 0:42:03 | 0:42:06 | |
you drink a lot, you are reasonably drunk now, OK? | 0:42:06 | 0:42:10 | |
And you will go home and do the same thing again, | 0:42:10 | 0:42:12 | |
and the car safely brings you home. | 0:42:12 | 0:42:14 | |
That's going to happen in the next five years. | 0:42:14 | 0:42:16 | |
So the future's bright for pizza-eating alcoholics, | 0:42:16 | 0:42:19 | |
so long as they live in the | 0:42:19 | 0:42:20 | |
operational designed domain of their vehicle. | 0:42:20 | 0:42:22 | |
But within these zones, level four does offer complete autonomy, | 0:42:24 | 0:42:29 | |
with no driving controls available to passengers at all. | 0:42:29 | 0:42:32 | |
However unlikely this vision of the future may seem, | 0:42:34 | 0:42:37 | |
this is exactly what some | 0:42:37 | 0:42:39 | |
manufacturers are promising us is just around the corner. | 0:42:39 | 0:42:43 | |
We're announcing Ford's intent to have a high-volume, SAE, level four, | 0:42:43 | 0:42:49 | |
fully autonomous vehicle in | 0:42:49 | 0:42:52 | |
commercial operation in 2021 in a ride-hailing | 0:42:52 | 0:42:57 | |
or ride-sharing service. | 0:42:57 | 0:42:59 | |
It's a bold claim, | 0:43:00 | 0:43:02 | |
and one that will test the navigational systems to their limit. | 0:43:02 | 0:43:06 | |
But even if these machines come good, | 0:43:06 | 0:43:08 | |
there's another subtle aspect to | 0:43:08 | 0:43:10 | |
driving that is sometimes overlooked. | 0:43:10 | 0:43:13 | |
Probably because it's such a human issue. | 0:43:13 | 0:43:15 | |
This small autonomous machine is called Jack Rabot. | 0:43:22 | 0:43:26 | |
By going about his smartly dressed business, | 0:43:27 | 0:43:30 | |
he's finding out how people behave around each other. | 0:43:30 | 0:43:32 | |
Jack Rabot learns from the behaviour of other people how to move around. | 0:43:35 | 0:43:40 | |
The more he looks at people, the better he will be in his navigation. | 0:43:40 | 0:43:44 | |
It turns out that we humans, when we navigate in crowded scene, | 0:43:45 | 0:43:49 | |
we read each other's behaviour, body language to avoid each other. | 0:43:49 | 0:43:53 | |
We respect personal space, yield right of way, and the culture, | 0:43:53 | 0:43:58 | |
the way people decide to move around is different, | 0:43:58 | 0:44:02 | |
because we all have different social behaviour, | 0:44:02 | 0:44:06 | |
and this behaviour can only be learned from the observation, | 0:44:06 | 0:44:09 | |
from the data. We cannot define them. | 0:44:09 | 0:44:12 | |
We cannot | 0:44:12 | 0:44:14 | |
write the rules that should be | 0:44:14 | 0:44:15 | |
applied in the UK and in Japan on the same time. | 0:44:15 | 0:44:19 | |
That's because, however adept Jack | 0:44:19 | 0:44:21 | |
might be at schmoozing with Californian students, | 0:44:21 | 0:44:24 | |
abandon him in, say, Dagenham, and he might struggle. | 0:44:24 | 0:44:28 | |
It's the same with driving. | 0:44:28 | 0:44:30 | |
The reason is that | 0:44:30 | 0:44:33 | |
the way Americans drive is not the | 0:44:33 | 0:44:35 | |
same as French and British people drive. | 0:44:35 | 0:44:39 | |
The only limiting factor of a level | 0:44:42 | 0:44:44 | |
four car is that it will only work in predefined areas. | 0:44:44 | 0:44:48 | |
A level five vehicle will be able to work anywhere, | 0:44:49 | 0:44:53 | |
whatever the driving conditions and | 0:44:53 | 0:44:55 | |
whatever the cultural driving conventions. | 0:44:55 | 0:44:58 | |
It might be, then, | 0:44:58 | 0:45:00 | |
that Jack is helping to deliver the ultimate driverless nirvana. | 0:45:00 | 0:45:04 | |
Level five. | 0:45:06 | 0:45:08 | |
We have now arrived at full autonomy. | 0:45:08 | 0:45:11 | |
Now, according to the masterplan, | 0:45:11 | 0:45:13 | |
the car will have capabilities at least equal to a human driver in | 0:45:13 | 0:45:17 | |
every possible driving scenario. | 0:45:17 | 0:45:19 | |
What this actually means is what it says. | 0:45:22 | 0:45:25 | |
Pop in a postcode or, if you're more rugged, | 0:45:25 | 0:45:27 | |
GPS coordinates, and off you go, anywhere, | 0:45:27 | 0:45:31 | |
and do anything you like on the way. | 0:45:31 | 0:45:33 | |
At FIVE AI, | 0:45:35 | 0:45:37 | |
Teena is hoping to witness the reasonably priced car's first steps | 0:45:37 | 0:45:41 | |
on the path to this full level five autonomy. | 0:45:41 | 0:45:44 | |
OK, so this is basically the first day in the real world. | 0:45:45 | 0:45:47 | |
It is the first day in the real world, indeed. | 0:45:47 | 0:45:49 | |
Yeah. So we're at a test track here, | 0:45:49 | 0:45:51 | |
so we're not going to sort of hopefully damage anything. | 0:45:51 | 0:45:54 | |
Excellent. Shall we go, then? | 0:45:54 | 0:45:56 | |
Let's go. | 0:45:56 | 0:45:58 | |
So what I'm going to do here... We're going to just hit return. | 0:45:58 | 0:46:01 | |
OK. | 0:46:01 | 0:46:02 | |
We're now off. I've got overall | 0:46:03 | 0:46:05 | |
control of the car with this dead man's | 0:46:05 | 0:46:07 | |
handle here, so if there's anything that goes disastrously wrong, | 0:46:07 | 0:46:10 | |
we can always stop. But as you can see, we're not going very fast, | 0:46:10 | 0:46:14 | |
actually, we're going about four or five miles an hour. | 0:46:14 | 0:46:16 | |
So it's... There's not a huge danger | 0:46:16 | 0:46:18 | |
of anything particularly going wrong. | 0:46:18 | 0:46:20 | |
Within one or two days, we'll be | 0:46:22 | 0:46:23 | |
going round much more complex tracks, | 0:46:23 | 0:46:26 | |
and within a few weeks, | 0:46:26 | 0:46:27 | |
we'll be able to deal with simple kind of junctions and certainly able | 0:46:27 | 0:46:32 | |
to deal with obstacles. | 0:46:32 | 0:46:33 | |
Actually turned out to be OK. Looked like it was avoiding those cones. | 0:46:34 | 0:46:38 | |
It did look like it was avoiding those cones. | 0:46:38 | 0:46:40 | |
Although we weren't fully sure. | 0:46:40 | 0:46:42 | |
We weren't fully sure. | 0:46:42 | 0:46:43 | |
It's early days, | 0:46:45 | 0:46:46 | |
but this particular electric car has got a fair way to go if it's to | 0:46:46 | 0:46:50 | |
achieve level five autonomy. | 0:46:50 | 0:46:51 | |
But if it does, it could revolutionise our driving world. | 0:46:55 | 0:46:59 | |
I'm excited by the prospect of driverless cars. | 0:47:02 | 0:47:04 | |
You'll just sit in the cab and it will take you where you want to go. | 0:47:04 | 0:47:07 | |
And to me, that's good. I can do other things. | 0:47:07 | 0:47:10 | |
Sure, it'd be a big deal to be able to get into your car and curl up in | 0:47:10 | 0:47:12 | |
the back seat and have a nap while you go from A to B, but | 0:47:12 | 0:47:16 | |
if you think about how cities and whole countries are built around the | 0:47:16 | 0:47:19 | |
road network, | 0:47:19 | 0:47:21 | |
the changes that could happen to that are enormous. | 0:47:21 | 0:47:24 | |
And that's really the challenge - | 0:47:24 | 0:47:27 | |
what does the infrastructure need to look like for these vehicles? | 0:47:27 | 0:47:30 | |
One major challenge is that, for the past century or so, | 0:47:36 | 0:47:39 | |
we've been building our world around a totally different kind of car. | 0:47:39 | 0:47:43 | |
And before you even think about autonomy, | 0:47:43 | 0:47:46 | |
you need to gear up for electric. | 0:47:46 | 0:47:48 | |
My name's Gareth Dunsmore. | 0:47:50 | 0:47:52 | |
I run electric vehicles for Nissan in Europe. | 0:47:52 | 0:47:54 | |
I perhaps enjoy driving | 0:47:54 | 0:47:56 | |
a Leaf more than any vehicle I've driven, | 0:47:56 | 0:47:58 | |
an electric vehicle more than any other vehicle I've driven. | 0:47:58 | 0:48:01 | |
-Really? -Yeah. -You just come here, you have to say that, don't you? | 0:48:01 | 0:48:03 | |
No. It's different, it's a different driving experience. | 0:48:03 | 0:48:06 | |
If you've not done it you can't explain it, but it's... | 0:48:06 | 0:48:09 | |
It's instant acceleration, | 0:48:09 | 0:48:10 | |
and that... You don't just... You don't get that even in a GTR. | 0:48:10 | 0:48:14 | |
Apart from the obvious thrill of | 0:48:14 | 0:48:15 | |
driving Nissan's entry-level electric | 0:48:15 | 0:48:18 | |
option, Gareth has another reason to be evangelical | 0:48:18 | 0:48:21 | |
about alternative power. | 0:48:21 | 0:48:23 | |
Vehicles are far easier to automate if they are also electric. | 0:48:23 | 0:48:27 | |
On top of that, though, I think there's a broader point. | 0:48:27 | 0:48:30 | |
Looking at cities and looking at what we're trying to achieve, | 0:48:30 | 0:48:33 | |
it's about moving to zero emissions and zero fatalities. | 0:48:33 | 0:48:37 | |
So combining the two technologies together makes | 0:48:37 | 0:48:40 | |
more sense from a customer perspective, | 0:48:40 | 0:48:43 | |
to be able to bring forward both at the same time. | 0:48:43 | 0:48:45 | |
In planning for this electric utopia, | 0:48:47 | 0:48:49 | |
Gareth teamed up with British architects Foster + Partners. | 0:48:49 | 0:48:53 | |
Together, they came up with a | 0:48:53 | 0:48:55 | |
version of the future that imagines how | 0:48:55 | 0:48:57 | |
electric autonomous cars might do more than just drive themselves. | 0:48:57 | 0:49:01 | |
I think the challenge | 0:49:03 | 0:49:05 | |
is integrating it into, you know, existing city fabric. | 0:49:05 | 0:49:08 | |
And that's where the conversations with Nissan about taking one part of | 0:49:08 | 0:49:11 | |
it and the charging strategy and autonomously doing that... | 0:49:11 | 0:49:14 | |
They're all going to have their own nuances which need to be solved. | 0:49:14 | 0:49:17 | |
I think it's shifted our perspective slightly. | 0:49:17 | 0:49:19 | |
We kind of, you know...probably would have tackled the problem in a | 0:49:19 | 0:49:23 | |
very, very different way. But you realise | 0:49:23 | 0:49:25 | |
the issues facing urban development around the world, | 0:49:25 | 0:49:27 | |
you need a very integrated approach to transport, in particular, | 0:49:27 | 0:49:31 | |
how you move from public to private transport, | 0:49:31 | 0:49:33 | |
how you cover big distances, what's the economics of all of that, | 0:49:33 | 0:49:37 | |
how do you make it work? | 0:49:37 | 0:49:39 | |
With Nissan's wireless charging and universal connectivity, | 0:49:39 | 0:49:43 | |
our vehicles could autonomously charge themselves | 0:49:43 | 0:49:47 | |
and then re-park so another vehicle | 0:49:47 | 0:49:49 | |
on the street could use the same bay, | 0:49:49 | 0:49:52 | |
all while you sleep. | 0:49:52 | 0:49:54 | |
I think the thing that we're quite excited about is | 0:49:54 | 0:49:57 | |
there's less reliance on having your car always parked outside your | 0:49:57 | 0:50:00 | |
house. Once we release them, your car can come from round the corner. | 0:50:00 | 0:50:04 | |
We can start opening up some of these sort of residential areas, | 0:50:04 | 0:50:07 | |
and they'll feel much, you know... | 0:50:07 | 0:50:09 | |
It almost goes back to, you know, a hundred years ago, streets | 0:50:09 | 0:50:12 | |
become actually a really nice place to be outside, | 0:50:12 | 0:50:14 | |
not in your little sort of hidden gardens. | 0:50:14 | 0:50:17 | |
And we become more sociable again and | 0:50:17 | 0:50:18 | |
use these spaces that are more shared. | 0:50:18 | 0:50:20 | |
Self-driving cars are going to have a huge impact. | 0:50:22 | 0:50:25 | |
If all cars are self-driving, we can get rid of the streetlights, | 0:50:25 | 0:50:28 | |
cos they won't need them, they'll be able to navigate through complex | 0:50:28 | 0:50:31 | |
intersections with no collisions. | 0:50:31 | 0:50:33 | |
Right now, we have in the United States 100 million cars. | 0:50:34 | 0:50:37 | |
They are parked 97% of the time, only driven 3%. | 0:50:37 | 0:50:41 | |
So in the future, we're going to have less traffic, | 0:50:41 | 0:50:44 | |
and once we have robotic, self-driving car taxi services, | 0:50:44 | 0:50:48 | |
we don't need parking spaces any more. | 0:50:48 | 0:50:50 | |
It means that the cities will look nicer. | 0:50:50 | 0:50:52 | |
If cars are not crashing, we're not going to need as many doctors, | 0:50:52 | 0:50:56 | |
and we can totally reimagine the car. | 0:50:56 | 0:50:59 | |
We don't need the heavy steel, rigid bodies of these cars to protect the | 0:50:59 | 0:51:03 | |
inhabitants. They could be made of more environmental friendly, | 0:51:03 | 0:51:06 | |
more lightweight materials. | 0:51:06 | 0:51:08 | |
So great societal benefits with self-driving cars. | 0:51:08 | 0:51:12 | |
Well, that's settled, then. | 0:51:14 | 0:51:15 | |
Driverless cars will usher in a | 0:51:15 | 0:51:17 | |
world where road traffic accidents will be | 0:51:17 | 0:51:20 | |
a thing of the past, where the lion will lie down with the lamb, | 0:51:20 | 0:51:23 | |
swords will be beaten into ploughshares, | 0:51:23 | 0:51:26 | |
and people will be nicer, kinder, | 0:51:26 | 0:51:28 | |
happier and richer. | 0:51:28 | 0:51:30 | |
Except, of course, they won't. | 0:51:31 | 0:51:33 | |
I think a wholly autonomous, | 0:51:34 | 0:51:36 | |
driverless future could have more social impacts than we imagine. | 0:51:36 | 0:51:40 | |
I mean, first off, you've got to | 0:51:40 | 0:51:41 | |
think about all the people who drive to make a living. | 0:51:41 | 0:51:44 | |
And so you've got to think about how automation will affect | 0:51:44 | 0:51:46 | |
their livelihood. | 0:51:46 | 0:51:49 | |
There are about half a million taxi drivers, delivery drivers | 0:51:49 | 0:51:52 | |
and bus drivers in the UK, | 0:51:52 | 0:51:54 | |
who might well need to look for alternative employment. | 0:51:54 | 0:51:57 | |
But as if that wasn't bad enough, | 0:51:58 | 0:52:00 | |
driverless cars might even make our | 0:52:00 | 0:52:02 | |
cities' congestion problems even worse. | 0:52:02 | 0:52:04 | |
I can see very quickly a time where people won't actually stop their car | 0:52:05 | 0:52:09 | |
driving when they want to go to the shops. | 0:52:09 | 0:52:11 | |
They'll go to the shops, they'll get out, | 0:52:11 | 0:52:13 | |
and then they'll tell their car to drive around the block and wait till | 0:52:13 | 0:52:15 | |
they're ready. | 0:52:15 | 0:52:17 | |
And the problems don't stop there. | 0:52:17 | 0:52:20 | |
There are practical and ethical issues yet to be resolved. | 0:52:20 | 0:52:23 | |
It'll be difficult for autonomous cars to coexist with people driving | 0:52:24 | 0:52:27 | |
cars, which will obviously be the case for many years to come. | 0:52:27 | 0:52:31 | |
The cars with drivers are | 0:52:31 | 0:52:32 | |
potentially going to take advantage of the | 0:52:32 | 0:52:35 | |
driverless cars, which they know have to stop and give way when their | 0:52:35 | 0:52:38 | |
sensors detect that something's wrong. | 0:52:38 | 0:52:40 | |
What happens when it eventually encounters a no-win scenario, | 0:52:43 | 0:52:46 | |
when it actually has to have an accident, where it cannot avoid it? | 0:52:46 | 0:52:50 | |
If you get to the point where the car has to choose between | 0:52:50 | 0:52:56 | |
an old man in a vehicle or a young lady and an infant in the other car, | 0:52:56 | 0:53:01 | |
and it knows it's going to hit one of them, | 0:53:01 | 0:53:03 | |
these cars are going to have to be taught to make these moral, | 0:53:03 | 0:53:06 | |
ethical judgments that you make in a snap decision. | 0:53:06 | 0:53:10 | |
Some car manufacturers have already said that they're going to teach the | 0:53:10 | 0:53:13 | |
car to favour the driver above all else. | 0:53:13 | 0:53:17 | |
Now, what if you don't want that? | 0:53:17 | 0:53:19 | |
You might actually set those parameters in the vehicle yourself, | 0:53:19 | 0:53:22 | |
the level of morality for the car. | 0:53:22 | 0:53:24 | |
Where some driver-assist technologies have been introduced, | 0:53:27 | 0:53:30 | |
there have already been problems. | 0:53:30 | 0:53:33 | |
A Tesla crashed, killing its driver, | 0:53:33 | 0:53:35 | |
and Uber suspended their fleet of | 0:53:35 | 0:53:37 | |
semi-autonomous test cars following a non-fatal collision. | 0:53:37 | 0:53:42 | |
But despite this, the technology moves forward. | 0:53:42 | 0:53:45 | |
It's three weeks since Teena was at FIVE AI, and today, | 0:53:53 | 0:53:56 | |
she's going back to meet them for the last time, | 0:53:56 | 0:53:58 | |
to see how their car's getting on | 0:53:58 | 0:54:00 | |
with its assault on level five autonomy. | 0:54:00 | 0:54:02 | |
What we're now going to do is just engage our robot. | 0:54:06 | 0:54:09 | |
Press that button there. Robot is now in control. | 0:54:09 | 0:54:12 | |
And off we go. | 0:54:12 | 0:54:14 | |
OK, so the car's now setting off. | 0:54:14 | 0:54:17 | |
See it's quickly got up to... | 0:54:17 | 0:54:19 | |
Doing about 12 miles an hour on this access road here. | 0:54:19 | 0:54:21 | |
So right now, we're driving entirely autonomous. | 0:54:21 | 0:54:23 | |
We're driving entirely autonomously. | 0:54:23 | 0:54:25 | |
Actually, this robot's getting a bit rattly. | 0:54:25 | 0:54:28 | |
-It is. -A few of the things we've | 0:54:28 | 0:54:30 | |
really improved over the last few weeks | 0:54:30 | 0:54:32 | |
is we've managed to sort out a lot of the visual odometry, | 0:54:32 | 0:54:35 | |
so the car can work out really | 0:54:35 | 0:54:36 | |
accurately where it is in three-dimensional | 0:54:36 | 0:54:39 | |
space. Just from the cameras and the pixels on the camera and their | 0:54:39 | 0:54:43 | |
-motion... -Yeah. -..it can actually work out... | 0:54:43 | 0:54:45 | |
It can in fact go round really quite long tracks, | 0:54:45 | 0:54:48 | |
and within a few centimetres can detect exactly where it is in | 0:54:48 | 0:54:51 | |
three-dimensional space without the use of a map. | 0:54:51 | 0:54:54 | |
-Yeah. -So that's pretty cool. | 0:54:54 | 0:54:56 | |
Although, not quite cool enough. | 0:54:56 | 0:54:59 | |
An old-fashioned level zero | 0:54:59 | 0:55:01 | |
interception is required to save the car from itself. | 0:55:01 | 0:55:04 | |
Whoa, it's too close. | 0:55:04 | 0:55:06 | |
Let's just make sure this thing is not going to hit that barrier there. | 0:55:06 | 0:55:10 | |
There's some latencies in the system that we're now going to iron out. | 0:55:10 | 0:55:13 | |
And when you say latency, what's actually happening | 0:55:13 | 0:55:15 | |
is the car's not responding as fast as you'd like. | 0:55:15 | 0:55:17 | |
That's right. Yeah, it takes... | 0:55:17 | 0:55:19 | |
It takes a few hundreds of | 0:55:19 | 0:55:21 | |
milliseconds to go all the way through from | 0:55:21 | 0:55:23 | |
seeing data in the cameras to | 0:55:23 | 0:55:25 | |
actually a decision that actually controls the wheel. | 0:55:25 | 0:55:29 | |
And that gap between the two | 0:55:29 | 0:55:31 | |
means that by the time we actually apply control, | 0:55:31 | 0:55:34 | |
it's to a situation that was true maybe half a second ago. | 0:55:34 | 0:55:37 | |
-Right. -And, you know, | 0:55:37 | 0:55:39 | |
we need to take all that into | 0:55:39 | 0:55:40 | |
account in the way we design the car. | 0:55:40 | 0:55:42 | |
Also the way we speed up some of the algorithms, as well. | 0:55:42 | 0:55:44 | |
-Yeah. -Stop it there at that point, actually. | 0:55:44 | 0:55:47 | |
We're about to hit the roundabout. So just go round this roundabout. | 0:55:47 | 0:55:49 | |
Basically, it is still in a slightly drunken manner, but... | 0:55:49 | 0:55:53 | |
Yeah, it is... It is suffering from a serious case of PIO. | 0:55:53 | 0:55:56 | |
Pilot-induced oscillation. | 0:55:56 | 0:55:58 | |
Responding to the signal too late. | 0:55:58 | 0:56:00 | |
Very well documented in lots of aeronautics texts. | 0:56:00 | 0:56:02 | |
Well, there you go. You know, | 0:56:02 | 0:56:04 | |
this is why we should have cross-disciplinary... | 0:56:04 | 0:56:06 | |
Yeah, you need some dynamicists on board. | 0:56:06 | 0:56:08 | |
We do need some dynamicists, I tell you. Yeah. | 0:56:08 | 0:56:12 | |
Despite the obvious challenges, Stan remains optimistic. | 0:56:12 | 0:56:15 | |
In 2019, we're going to be driving autonomously in urban scenes, | 0:56:18 | 0:56:22 | |
and I reckon there's going to be commercial, you know, | 0:56:22 | 0:56:25 | |
within one-and-a-half to two years after that. | 0:56:25 | 0:56:27 | |
FIVE AI have chosen the hardest problem to solve, | 0:56:28 | 0:56:32 | |
that of complete autonomy, | 0:56:32 | 0:56:34 | |
so their journey in reaching that goal could well be a long one. | 0:56:34 | 0:56:37 | |
The prospect of replacing our cars | 0:56:44 | 0:56:46 | |
with driverless ones tends to split opinion. | 0:56:46 | 0:56:49 | |
You know, we've done sort of surveys | 0:56:51 | 0:56:52 | |
and focus groups and things like that | 0:56:52 | 0:56:55 | |
with people, and it's honestly 50-50. | 0:56:55 | 0:56:58 | |
Some people hate the idea of | 0:56:58 | 0:57:00 | |
relinquishing the pleasure of driving. | 0:57:00 | 0:57:02 | |
I love driving. I'm an automotive journalist, | 0:57:02 | 0:57:05 | |
so I have to love driving or I wouldn't do my job. | 0:57:05 | 0:57:07 | |
Others relish the prospect of | 0:57:07 | 0:57:09 | |
freeing up wasted time behind the wheel. | 0:57:09 | 0:57:12 | |
If it can make my commute into work that much easier, | 0:57:12 | 0:57:15 | |
I won't have to stress about it in the morning, that'd be grand. | 0:57:15 | 0:57:18 | |
If I could just sit back and read a book, listen to some music, | 0:57:18 | 0:57:21 | |
catch up on some sleep, that would be great. | 0:57:21 | 0:57:24 | |
For some, it's the ultimate freedom. | 0:57:24 | 0:57:26 | |
For others, it's an opportunity for less freedom, for a longer work day. | 0:57:26 | 0:57:30 | |
All those times when you were travelling, | 0:57:32 | 0:57:34 | |
and it was a break from the phone calls and it was a break from the | 0:57:34 | 0:57:38 | |
e-mail, I think those times would be freed up for us to actually carry on | 0:57:38 | 0:57:43 | |
working and the daily grind. | 0:57:43 | 0:57:46 | |
But despite the reservations, the die seems to have been cast. | 0:57:47 | 0:57:52 | |
The dawn of the driverless car is here. | 0:57:52 | 0:57:55 | |
The challenge for the technologists is to make sure that the transition | 0:57:55 | 0:57:59 | |
into reality is as beguilingly smooth as the PR that surrounds it. | 0:57:59 | 0:58:04 | |
I can't foresee the future, but I can build it. | 0:58:04 | 0:58:08 | |
The reason that I believe it's good to be a technology optimist is | 0:58:08 | 0:58:11 | |
throughout the entire history of the human race, | 0:58:11 | 0:58:13 | |
technology has empowered us. | 0:58:13 | 0:58:15 | |
From the very early days, the Bronze Age, the Stone Age, | 0:58:15 | 0:58:19 | |
to the day of the smartphones and modern medicine, it has freed us, | 0:58:19 | 0:58:23 | |
it has levelled the playing field for everybody, | 0:58:23 | 0:58:25 | |
and has empowered us as a human race. | 0:58:25 | 0:58:27 | |
Why stop that? | 0:58:27 | 0:58:28 |