Name: The $800M Robo Taxi That Could Beat Uber
Uploaded: 2018-09-21T03:41:49.000Z
Duration: 9 min 1 s
Description: Thousands of YouTube videos with English-Chinese subtitles! Now you can learn to understand native speakers, expand your vocabulary, and improve your pronunciation...

They're the founders of a self driving car company called Zoox

and they've decided it makes sense to try and compete against this

Waymo the Google self driving car project, Cruise Automation, Daimler, Baidu, Uber, and Lyft

But Zoox is different. its rivals are adding self driving technology to existing cars

Meanwhile Zoox is making a new type of vehicle

and its early incarnation. Looks like the love child of the Matrix, Mad Max, and Transformers franchises

He hails from Australia and Down Under he was a well-known designer and ad guy

One day he had an epiphany. A vision of the future full of robotic vehicles and came here to build them

I sort of saw the future of how the whole system could work

I said ok I need to find a brilliant computer scientist

so that led me on a journey for six months actually where I ended up finding Jesse

That brilliant computer scientist is Dr. Jesse Levinson

A Stanford grad who was sought after by none other than Google itself

I actually visited a Google X in 2013 to come and speak on my vision for autonomy

In talking a fellow he there's one guy and he's really good and we can't get him

a lot of your colleagues have run off to companies like Google or or Uber to retrofit their cars

I've been working on some driving cars since 2005 and

it's pretty clear even back then that there's going to be one of the more transformative technologies for society

It wasn't enough to just solve the technology

You need a product and you need a business that makes sense

And I hadn't seen that in any other company until I met Tim and he shared his high level vision

And that vision is a system of completely autonomous robotic vehicles

that can be summoned just like a Lyft, or dare I saw it, an Uber

which is that AI and mobility will take us out of the age of the automobile

and into the next mobility age which for us is robotics autonomous transportation

Tim and Jesse take me to the Alameda Naval Air Station

where I'm about to get a test drive and see how a Zooks robot handles

these are very unique vehicles. They're designed in a way that would only make sense

The power trains are designed for full autonomy they are fully redundant and they have joule high voltage

batteries they have joule low voltage architectures

you can see that rigged with computer sensors. They can drive without any human intervention

You want to go for a ride?. Yeah I do. All right. Jump in

Zoox has placed cones on the tarmac. The robot is going to drive out and try to dodge them

And I get to do the whole test sitting backwards because Tim thought it would be funny

and with one swift keystroke I give complete control of my life over to an ai

These AI robots crank through the course at forty five miles per hour.

And the VH5 is set to rip through it at a delightful seventy five miles per hour

We can choose all wheel steering, two wheel steering, crab steering,

Zoox got its start here in this abandoned firehouse back in 2015 with only six employees

They've since raised an astonishing eight hundred million dollars

and have expanded their operations into a one stop robot making shop

Zoox now on their way to turning these prototypes into full fledged vehicles by 2020

For the moment though Zoox does most of its software testing in these

Modified versions of Toyota Highlander's which have the decided advantage of being street legal.

We're actually going to get you to drive autonomously from this facility to our new headquarters in Foster City

And then Jesse is going to jump in and you can travel to downtown San Francisco

We don't know any company in the world that is driving that multi-dimensional degree of difficulty

That was a cyclist yeah and that's the car behind us

Is something like the speed limit, does that get taken by the cameras seeing the signs

or is that something that has to get entered in as you map the world?

If there's a sign that's been vandalized or it's missing, or you know, it's got growth on it

The vehicle wouldn't know what the speed limit is and that could cause an incident right

So speed limits are really well understood on the road networks. So we put that in as a primer on the map

So coming up right here we have to do a left turn yield

Yes so were we would be concerned about this car that's coming towards us from a prediction level

So we just want to make sure that they do the right thing

And now we're yielding so if you look on this screen see this yellow here

That means we're yielding for that vehicle

We're about to drive onto the freeway so this will be interesting

you can see there's quite a bit of traffic on our left

and the vehicles indicating at the moment

it's going to break a little bit harder because we've got this vehicle right on our 6

and we're running out of runway. And now we merge in.

So that would have been a hard situation even for a human driver.

We needed to merge and that person on the left was not letting us in

We tried to go in front. They stayed with us and then we had to break,

let them go past, and then we pulled in. That was, that was all autonomous

So one of the hardest things to do with autonomous driving is actually prediction

because you not only need to understand whats the state of the world is,

you have to say well what's the state of the world in the near future

And so that creates very quickly a large search tree and if the vehicle is going to drive with good performance

Subtitles ListPlay Video

The $800M Robo Taxi That Could Beat Uber

sense

brilliant

cruise

autonomy