Aicha Evans: Your self-driving robotaxi is almost here | TED

譯者: 麗玲 辛 審譯者: Shelley Tsang 曾雯海

I’m Aicha Evans,

我是艾查‧ 埃文斯，

I am from Senegal, West Africa,

來自西非的塞內加爾，

and I fell in love with technology, science and engineering

我很小的時候就愛上 科技、科學與工程。

at a very young age.

Three things happened.

接著發生了三件事。

I was studying in Paris,

我在巴黎就學，

and starting at seven years old,

七歲時開始搭飛機

flying back and forth between Dakar, Senegal and Paris

在達卡（塞內加爾首都） 與巴黎間來回，

as an unaccompanied minor.

一個未成年人獨自旅行。

So it wasn't just about the travel.

這不只是旅程奔波，

It was really about a portal to knowledge,

更是走進知識的大門、

different environments

身處不同的環境、

and adapting.

適應生活。

Second thing that happened

第二件發生的事，

was every time I was at home in Senegal,

每次我回到塞內加爾，

I wanted to talk to my friends in Paris.

我想跟我巴黎的朋友講電話，

So my dad got tired of the long-distance bills,

但我爸爸不想一直接到長途電話帳單，

so he put a little lock on the phone --

所以他把電話鎖上，

the rotary phone.

那種有撥號盤的老式電話。

I said, OK, no problem,

我說，好吧，沒問題，

hacked it,

想辦法破解，

and he kept getting the bills.

所以他還是收到帳單。

Sorry again, Dad, if you’re watching this someday.

爸爸，如果你看到這個演講，對不起。

And then, obviously, the internet was also emerging.

然後，第三件事，網路開始盛行。

So what really happened was that, in terms of technology,

我認為真正的影響在於

科技形塑了人們的經驗、 理解世界的方式，

I really saw it as something that shaped your experiences,

how you understand the world

and wanting to be part of it.

也讓人想參與其中。

And for me,

就我本身，

the common thread is that physical and virtual transportation --

這三件事的共同點是 實體與虛擬的交通工具，

because that’s really what that rotary phone was for me --

那台撥號電話對我就是交通工具，

are at the center of the innovation flywheel.

而這是創新進步的核心。

Now, fast-forward.

快轉到現在，

I’m here today,

我現在成為一個運動、產業的一份子，

I’m part of a movement and an industry

that is working on bringing transportation and technology together.

努力結合交通與科技。

Huh.

嗯，

It’s not just about your commutes.

我要談的不只是通勤方式，

It’s really about changing everything

而是所有的改變，

in terms of how we move people, goods and services, eventually.

如何移動人們、貨物、 終而改變服務方式，

That transformation involves robotaxis.

而這徹底轉變與自駕計程車有關。

Driverless cars again, really?

又是自駕車？

Yeah, yeah, yeah, I’ve heard it before.

對啦對啦，這我聽過了。

And by the way, they are always coming the next decade,

而且，你們總是說 未來十年內會上市，

and oh, by the way,

啊，還有，

there’s an alphabet soup of companies working on it

一堆字母代號的公司在做自駕車，

and we can’t even remember who’s who and who’s doing what.

我都搞不清楚誰是誰、在做什麼了。

Yeah?

是吧？

Audience: Yeah.

（觀眾：是。）

AE: Yeah, OK, well, this is not about personal, self-driving cars.

好的，不過，我不是要談 私人的自駕車。

Sorry to disappoint you.

抱歉讓你們失望了。

This is really about a few things.

我其實要談幾件事。

First of all,

首先，

personally and individually owned cars are a wasteful expense,

自用車輛很花錢也很浪費，

and they contribute to, basically, a lot of pollution

而且基本上造成很多污染，

and also traffic in urban areas.

還有城市裏的交通問題。

Second of all, there’s this notion of self-driving shuttles,

第二，有人提出自駕接駁車的概念，

but frankly, they are optimized for many.

但坦白說，接駁車要接送許多人，

They can’t take you specifically from point A to point B.

因此無法專門載你從Ａ點到Ｂ點。

OK, now we have --

好，現在我們有 —

hm, how am I going to say this --

嗯，怎麼說呢？—

the so-called “personal, self-driving” cars of today.

我們有所謂的「個人自駕車」。

Well, the reality is that those cars still require a human behind the wheel.

現實狀況是這些車輛仍然 需要一個人坐在駕駛座，

A safety driver.

一位安全駕駛。

Make no mistake about it.

當然，我也有一輛這種車，

I own one of those,

and when I’m in it,

當我坐在裏面，

I am a safety driver.

我就是那個安全駕駛。

So the question now becomes, What do we do with this?

現在的問題是， 我們怎麼解決這些狀況？

Well, we think that robotaxis,

首先，我們認為自駕計程車

first of all, they will take you specifically from point A to point B.

能夠載你從Ａ點到Ｂ點。

Second of all, when you're not using them,

第二、當你不用車的時候，

somebody else will be using them.

別人可以用。

And they are being tested today.

而且這些車已經在測試中。

When I say that we’re on the cusp of finally delivering that vision,

當我說我們即將看到這個願景成真，

there's actually reason to believe it.

這是有根據的。

At the core of self-driving technology is computer vision.

自駕科技的核心是電腦視覺。

Computer vision is a real-time representation,

電腦視覺能夠即時、

數位地呈現環境及環境內的互動。

digital representation, of the world and the interactions within it.

It has benefited from leaps and bounds of advancements

這些科技獲益於許多領域 日新月異的進步，

thanks to computer, sensors, machine learning and software innovation.

如電腦運算、感測器、 機器學習以及軟體創新。

At the core of computer vision are camera systems.

電腦視覺的核心是攝影系統。

Cameras basically help you see agents such as cars,

攝影機幫助你看到物體如車輛、

their locations and their actions,

它們的位置、動向、

pedestrians,

行人、他們的位置、動向、及手勢。

their locations,

their actions and their gestures.

In addition, there's also been a lot of advancements.

還有，電腦視覺已經進步很多。

So one example is our vehicle can see the skeleton framework

一個例子是我們的車輛現在 可以看到人的骨架，

to show you the direction of travel;

所以知道人的行走方向，

also to give you details, like, are you dealing with a construction worker

也能看到一些細節，例如， 在建築工地的工人，

in a construction zone

or are you dealing with a pedestrian that’s probably distracted

或者辨識行人分心看著手機的狀況。

because they are looking on their phone?

Now the reality, though --

不過現實是 —

and this is where it gets interesting --

有趣的部份來了－

is that the camera and the algorithms that help us really cannot yet match

欇影機以及演算法幫助我們看， 但還跟不上

the human brain’s ability to understand and interpret the environment.

人腦理解、詮釋環境的能力。

They just can’t.

它們還做不到。

Even though they provide you really high-resolution imaging

即便欇影機以及演算法 提供我們高畫質的影像，

that really gives you continuous coverage,

不間斷地處理，

that doesn’t get fatigued, impaired

不會累、不會削弱，

or, you know, drunk or anything like that,

或者喝醉等等，

at the end of the day,

總體來說，

there are still things that they can’t see and they can’t measure.

還有許多機器看不到、 無法衡量的部份。

So if we want autonomous-driving robotaxis soon,

因此，想趕快讓自駕計程車上路，

we have to supplement cameras.

我們必須補加攝影機。

Let me walk through some examples.

容我提幾個例子。

So radar gives you the direction of travel

雷達顯示物體的行經方向、

and measures the agent’s movement within centimeters per second.

測量物體每秒幾公分的移動。

Lidar gives you objects and shapes in the real world using depth perception

光學電達則顯示真實世界中物體 及其形狀，利用深度感知、

as well as long-range and the all-important night vision.

長程視力、以及非常重要的夜視能力。

And let me tell you about this,

我提這點，

because this is important to me personally and people who look like me.

因為這對我個人以及 我的種族來說很重要。

Then you have, also, long-wave infrared

同時，必須有長波紅外線，

where you are able to see agents that are emitting heat,

這樣車子就能看得到發熱的物體，

such as animals and humans.

例如動物及人類。

And that’s again,

再強調一次，

especially at night,

特別在晚上，

super important.

超級重要。

Now, every one of these sensors is very powerful by itself,

這些感測器每個都很厲害，

but when you put them together is when the magic happens.

當你把它們組合起來， 神奇的事發生了。

If you see with this vehicle, for example,

例如，你看這輛車，

you have these multiple sensor modalities

它有多重感測器模組

at all top four corners of the vehicle

裝在車輛上方的四個角落，

that basically provide you a 360-degree field of vision,

提供 360 度視野，

continuously,

不間斷的大量資訊，

in a redundant manner,

so that we don't miss anything.

這樣我們就不會有所遺漏。

And this is that same thing

各種資訊輸出結合在一起時， 也是相同情形。

with all of the different outputs fused together.

And looking at this, basically,

我們基本上在研究人們如何看、 如何處理資訊、如何學習，

and looking at what we see and how we are able to process the data,

then learn,

then continue to improve our driving,

然後持續改善自駕車的駕駛技術，

is what tells us that we have confidence,

這些歷程讓我們有信心說

this is the right approach

這是對的方向，

and this time it’s actually coming.

而且這次確實會成功。

Now, this is not, by the way, a brand new concept, OK?

順道一提，這並不是個新概念，對吧？

Humans have been basically using vision systems

人類已經使用視覺系統

to assist them for a long time.

做為輔助很長一段時間了。

Let me back up the boat a little bit,

容我倒回去一點，

because I know there’s a question that everybody’s asking,

因為我知道每個人都想問，

which is, “Hey, how are you going to deal with all the scenarios

「那你要怎麼處理街道上 所有的情況啊？」

out there on the streets today?”

Most of us are drivers,

我們都開車，

and it’s complicated out there.

知道街道路況很複雜。

Well, the truth is that there will always be edge scenarios

事實上總有一些邊緣案例

that sit at the boundary of our real-world testing

不在我們實測的範圍裏，

or that are just too dangerous to test on real streets.

或者因為太危險，無法 在街道上實際測試。

That is the truth,

這是事實，也會持續很長一段時間。

and it will be the truth for a very long time.

Human beings are pretty underrated in their abilities.

人類的能力其實相當被低估。

So what we do is we use simulation.

我們能做的就是模擬。

And with simulation,

藉由模擬，

we’re able to construct millions of scenarios

我們得以在虛構情境裏 建立數百萬個情況，

in a fabricated environment

so that we can see how our software would react.

看到軟體如何因應。

And that’s the simulation footage.

這就是模擬錄像。

You can see we’re building the world,

你可以看到我們建造這個世界，

we’re putting in scenarios

置入情境，

and we can add things,

我們可以加入物件， 移除物件，

remove things

and see how we would react.

檢視自駕車如何因應。

In addition, we have what's called a human in the loop.

再者，我們有所謂的訓練師 （負責引導 AI 系統學習的人）

This is very similar to aviation systems today.

這跟現在的飛行系統很像。

We don’t want the vehicle to get stuck,

我們不希望車輛卡住，

and there are rare times where it’s not going to know what to do.

雖然它很少遇到狀況 不知道如何處理，

So we have a team of teleguidance operators

我們還是設有遙導操作員團隊,

that are sitting at a control center,

坐在控制中心，

and if the vehicle knows that it’s going to be stuck

如果車輛知道它快要卡住，

or it doesn’t know what to do,

或者不知道該怎麼辦，

it asks for guidance and help

它會求助，請求指導，

and it receives it remotely

遠端接收指令後，

and then it proceeds.

它就可以繼續前行。

Now, none of these really are new concepts,

如我剛剛說的，這些方式 其實並不是新概念。

as I alluded to earlier.

Vision systems have been assisting humans for a long time,

視覺系統已經輔助人類一段時間了，

especially with things that are not visible to the naked eye.

特別是肉眼看不到的東西。

So ...

例如，

microscopes, right?

顯微鏡，對吧？

We’ve been studying microbes and cells for a long time.

我們已經研究微生物與細胞很久了。

Telescopes:

望遠鏡，

we’ve been studying and detecting galaxies millions of light-years away

我們已經研究、探測 好幾百萬光年以外的星系

for a long time.

好長一段時間了。

And both of these have caused us,

這兩種儀器已經幫助我們產業轉型，

for example,

to transform industries like medicine,

舉例來說，醫藥、農業、 天文物理學、還有其他。

farming,

astrophysics

and much more.

So when we talk about computer vision,

電腦視覺初步發展時，

when it started,

it was really a thought experiment

其實是個思想實驗，

to see if we could replicate what humans see using cameras.

檢驗我們是否可以用攝影機 複製人的視覺所見。

It has now graduated with sensors,

現在我們已經可以做到，

使用感測器、電腦、 人工智能、以及軟體創新，

computers,

AI

and software innovation

to be about surpassing what humans can see and perceive.

終而即將超越人類所見所感知。

We’ve made a lot of progress in this field,

在這個領域我們取得長足進步，

but at the end of the day,

但總括來說，

we have a lot more to do.

還有許多未竟之路。

And with an autonomous robotaxi,

我們希望自駕計程車安全無虞，

you want it to be safe,

right and reliable every single time,

每一次都正確、可靠，

which requires rigorous testing and optimization.

這需要嚴格的測試以及優化。

And when that happens

當我們達成這個要求，

and we reach that state,

we will wonder how we ever accepted

我們懷疑，人們是否還能接受或容忍

or tolerated

94 percent of crashes

94 % 交通事故 是由人類疏失造成的？

being caused by human [error].

So with computer vision,

有了電腦視覺的輔助，

we have the opportunity

我們有機會不只解決問題， 還能避免問題發生。

to move from problem-solving to problem-preventing.

And I truly, truly believe

我堅定相信

that the next generation of scientists and technologists

下一代的科學家與科技人員，

in, yes, Silicon Valley,

不只在矽谷，還有巴黎、

but in Paris,

in Senegal, West Africa

西非塞內加爾、全世界，

and all over the world,

will be exposed to computer vision applied broadly.

都能接觸到電腦視覺的廣泛應用。

And with that,

這樣，所有的產業都能轉型，

all industries will be transformed,

and we will experience the world in a different way.

而人們將以不同方式來體驗世界。

I hope you can join me in agreeing that this is a gift

我希望你跟我一樣認同

這個我們該給下一代的禮物即將到來，

that we almost owe our next generation that is coming,

because there are a lot of things that computer vision will help us solve.

因為電腦視覺可以幫我們 解決許多問題。

Thank you.

謝謝。

(Applause)

（掌聲）