Jennifer Healey: If cars could talk, accidents might be avoidable

Translator: Joseph Geni Reviewer: Morton Bast

Translator: Jure Mavrič Reviewer: Kaja Kren

Let's face it:

Priznajmo si:

Driving is dangerous.

Vožnja je nevarna.

It's one of the things that we don't like to think about,

Je ena od stvari o katerih ne razmišljamo radi,

but the fact that religious icons and good luck charms

ampak dejstvo, da se religiozne ikone in nalepke za srečo

show up on dashboards around the world

pojavljajo na armaturnih ploščah po svetu,

betrays the fact that we know this to be true.

priča o dejstvu, da se tega zavedamo.

Car accidents are the leading cause of death

Prometne nesreče so glavni povzročitelj smrti

in people ages 16 to 19 in the United States --

pri ljudeh od 16 do 19 let, v Združenih Državah --

leading cause of death --

glavni vzrok smrti --

and 75 percent of these accidents have nothing to do

in 75 procentov teh nesreč ni v povezavi

with drugs or alcohol.

z drogami ali alkoholom.

So what happens?

Torej, kaj se dogaja?

No one can say for sure, but I remember my first accident.

Nihče ne more zagotovo vedeti, ampak spomnim se svoje prve nesreče.

I was a young driver out on the highway,

Bila sem mladi voznik na avtocesti

and the car in front of me, I saw the brake lights go on.

in pri avtu pred mano sem videla, da so se prižgale zavorne luči.

I'm like, "Okay, all right, this guy is slowing down,

Pomislila sem "V redu, ta upočasnjuje,

I'll slow down too."

tudi samo bom upočasnila."

I step on the brake.

Pritisnila sem na zavoro.

But no, this guy isn't slowing down.

Ampak ne, ta pred mano ne upočasnjuje.

This guy is stopping, dead stop, dead stop on the highway.

Ta se ustavlja, popolnoma, ustavlja se sredi avtoceste.

It was just going 65 -- to zero?

Šel je iz 65 (100 km/h) -- do nič?

I slammed on the brakes.

Pritisnila sem na zavoro.

I felt the ABS kick in, and the car is still going,

Čutila sem, da se je uklopil ABS in avto je še kar peljal,

and it's not going to stop, and I know it's not going to stop,

in ne bo se ustavil, in vem, da se ne bo ustavil,

and the air bag deploys, the car is totaled,

in zračna blazina se je sprožila, avto je bil totalka.

and fortunately, no one was hurt.

Na srečo nihče ni bil poškodovan.

But I had no idea that car was stopping,

Ampak nisem vedela, da se bo tisti avto ustavil,

and I think we can do a lot better than that.

in mislim, da smo sposobni česa boljšega.

I think we can transform the driving experience

Mislim, da lahko preoblikujemo vožnjo,

by letting our cars talk to each other.

s tem, da dovolimo, da se naši avti pogovarjajo med sabo.

I just want you to think a little bit

Želim, da malo pomislite

about what the experience of driving is like now.

kako je dandanes voziti.

Get into your car. Close the door. You're in a glass bubble.

Sedeš v avto. Zapreš vrata. Si v steklenem mehurčku.

You can't really directly sense the world around you.

Ne moreš imeti neposrednega občutka sveta okoli tebe.

You're in this extended body.

Si v tem povečanem telesu.

You're tasked with navigating it down

Tvoja naloga je, da ga usmerjaš

partially-seen roadways,

po delno vidnih cestah,

in and amongst other metal giants, at super-human speeds.

med drugimi jeklenimi velikani, pri super-človeški hitrosti.

Okay? And all you have to guide you are your two eyes.

Ja? In vse kar imaš kot vodilo, sta tvoji dve očesi.

Okay, so that's all you have,

Torej, to je vse kar imaš,

eyes that weren't really designed for this task,

oči, ki niso bile ustvarjene za to nalogo.

but then people ask you to do things like,

Potem ljudje rečejo, stvari, kot so:

you want to make a lane change,

Hočeš zamenjati pas,

what's the first thing they ask you do?

kaj je prva stvar, ki ti jo naročijo?

Take your eyes off the road. That's right.

Nehaj gledati na cesto. Tako je prav.

Stop looking where you're going, turn,

Ne glej kam greš, obrni se,

check your blind spot,

preveri mrtve kote,

and drive down the road without looking where you're going.

in vozi po cesti, ne da bi gledal kam pelješ.

You and everyone else. This is the safe way to drive.

Ti in vsi ostali. To je varen način vožnje.

Why do we do this? Because we have to,

Zakaj to počnemo? Ker moramo.

we have to make a choice, do I look here or do I look here?

Moram sprejeti odločitev, naj gledam sem ali naj gledam sem.

What's more important?

Kaj je bolj pomembno?

And usually we do a fantastic job

In po navadi nam gre odlično

picking and choosing what we attend to on the road.

z izbiranjem in razmišljanjem čemu na cesti se posvetimo.

But occasionally we miss something.

Ampak včasih kaj spregledamo.

Occasionally we sense something wrong or too late.

Včasih kaj zaznamo narobe ali pa prepozno.

In countless accidents, the driver says,

V nešteto nesrečah, voznik pravi:

"I didn't see it coming."

"Nisem ga videl."

And I believe that. I believe that.

In temu verjamem. Temu verjamem.

We can only watch so much.

Vsega ne moremo videti.

But the technology exists now that can help us improve that.

Ampak danes imamo tehnologijo, ki nam lahko pomaga, da to izboljšamo.

In the future, with cars exchanging data with each other,

V prihodnosti, ko si bodo avti izmenjevali podatke,

we will be able to see not just three cars ahead

ne bomo videli le za tri avte naprej,

and three cars behind, to the right and left,

ampak za tri avte nazaj, v desno in levo,

all at the same time, bird's eye view,

vse to hkrati, iz ptičje perspektive,

we will actually be able to see into those cars.

pravzaprav bomo videli v tiste avte.

We will be able to see the velocity of the car in front of us,

Videli bomo lahko hitrost avta, ki je pred nami,

to see how fast that guy's going or stopping.

da vidimo kako hitro pospešuje ali se ustavlja.

If that guy's going down to zero, I'll know.

Če se popolnoma ustavlja, bom vedela.

And with computation and algorithms and predictive models,

In z izračuni in algoritmi in modeli predvidevanja,

we will be able to see the future.

bomo lahko videli v prihodnost.

You may think that's impossible.

Lahko se vam zdi nemogoče.

How can you predict the future? That's really hard.

Kako lahko predvidiš prihodnost? To je težko.

Actually, no. With cars, it's not impossible.

V bistvu ne. Z avti ni nemogoče.

Cars are three-dimensional objects

Avti so tridimenzionalni predmeti

that have a fixed position and velocity.

s stalnim položajem in hitrostjo.

They travel down roads.

Potujejo po cesti.

Often they travel on pre-published routes.

Pogosto potujejo po vnaprej zastavljeni poti.

It's really not that hard to make reasonable predictions

Res ni tako težko ustvariti razumnih predvidevanj

about where a car's going to be in the near future.

glede tega, kje bo avto v prihodnosti.

Even if, when you're in your car

Tudi, če si v avtu

and some motorcyclist comes -- bshoom! --

in pride mimo motorist --bshoom!--

85 miles an hour down, lane-splitting --

85 milj na uro (140 km/h).

I know you've had this experience --

Vem, da ste to že doživeli --

that guy didn't "just come out of nowhere."

ta motorist ni "kar prišel od nikoder".

That guy's been on the road probably for the last half hour.

Ta motorist je bil na cesti verjetno zadnje pol ure.

(Laughter)

(Smeh)

Right? I mean, somebody's seen him.

Res? Mislim, gotovo ga je nekdo videl.

Ten, 20, 30 miles back, someone's seen that guy,

10, 20, 30 milj nazaj, nekdo ga je videl

and as soon as one car sees that guy

in takoj, ko ga en avto vidi

and puts him on the map, he's on the map --

in ga postavi na zemljevid, je na zemljevidu --

position, velocity,

položaj, hitrost,

good estimate he'll continue going 85 miles an hour.

lahko rečemo, da bo pot nadaljeval pri 85 mph (140 km/h).

You'll know, because your car will know, because

Ti boš vedel, ker bo vedel tvoj avto, ker

that other car will have whispered something in his ear,

mu bo tisti drugi avto zašepetal na uho,

like, "By the way, five minutes,

npr.: "Mimogrede, pet minut,

motorcyclist, watch out."

motorist, pazi."

You can make reasonable predictions about how cars behave.

Lahko narediš razumna predvidevanja o vedenju avtov.

I mean, they're Newtonian objects.

Mislim, saj so 'Newtonski' predmeti.

That's very nice about them.

To je pri njih zelo prikladno.

So how do we get there?

Torej, kako do tega?

We can start with something as simple

Začnemo lahko z nečim preprostim,

as sharing our position data between cars,

kot je izmenjava podatkov o polažaju med avti,

just sharing GPS.

samo z delitvijo GPS-ja.

If I have a GPS and a camera in my car,

Če imam GPS in kamero v avtu,

I have a pretty precise idea of where I am

se mi kar zdi kje sem

and how fast I'm going.

in kako hitro peljem.

With computer vision, I can estimate where

Z računalniškim pogledom lahko predvidim kje

the cars around me are, sort of, and where they're going.

so avti okoli mene, nekako, in kam grejo.

And same with the other cars.

In enako je z ostalimi avti.

They can have a precise idea of where they are,

Lahko natančno vedo kje so,

and sort of a vague idea of where the other cars are.

in se jim zdi kje so ostali avti.

What happens if two cars share that data,

Kaj se zgodi, če si dva avta delita podatke,

if they talk to each other?

če se pogovarjata?

I can tell you exactly what happens.

Lahko vam povem natanko kaj se zgodi.

Both models improve.

Oba modela se izboljšata.

Everybody wins.

Vsi imajo korist.

Professor Bob Wang and his team

Profeso Bob Wang in njegova ekipa

have done computer simulations of what happens

so naredili računalniško simulacijo kaj se zgodi,

when fuzzy estimates combine, even in light traffic,

ko se približna predvidenja združi, celo pri malo prometa,

when cars just share GPS data,

kjer si avti le delijo GPS podatke.

and we've moved this research out of the computer simulation

In to raziskavo smo premaknili iz računalniške simulacije

and into robot test beds that have the actual sensors

v robotske testne neprave s senzorji,

that are in cars now on these robots:

ki so v avtih, zdaj na te robote:

stereo cameras, GPS,

stereo kamere, GPS

and the two-dimensional laser range finders

in dvodimenzionalne laserske pregledovalce območja,

that are common in backup systems.

ki so navadno v vzvratnih sistemih.

We also attach a discrete short-range communication radio,

Dodali smo tudi diskreten radio za komunikacijo na kratke razdalje,

and the robots talk to each other.

in roboti se med sabo pogovarjajo.

When these robots come at each other,

Ko se ti roboti približajo eden drugemu,

they track each other's position precisely,

si izmenjajo natančne podatke o položaju

and they can avoid each other.

in lahko se eden drugemu izogibajo.

We're now adding more and more robots into the mix,

Zdaj dodajamo več in več robotov

and we encountered some problems.

in odkrili smo nekaj problemov.

One of the problems, when you get too much chatter,

Eden od problemov se pojavi, ko je preveč tega pogovarjanja,

it's hard to process all the packets, so you have to prioritize,

potem je težko obdelati toliko podatkov, zato je treba nekatere prioritizirati

and that's where the predictive model helps you.

in s tem nam pomagajo modeli predvidevanja.

If your robot cars are all tracking the predicted trajectories,

Če vaši robotski avti sledijo predvidenim potem,

you don't pay as much attention to those packets.

se ni potrebno posvečati vsem njim.

You prioritize the one guy

Prednost daš tistemu,

who seems to be going a little off course.

za katerega se vidi, da mogoče ne bo sledil načrtu.

That guy could be a problem.

Ta zna biti problem.

And you can predict the new trajectory.

In lahko predvidiš novo pot.

So you don't only know that he's going off course, you know how.

Tako da ne veš samo da pelje izven načrta, ampak veš kako.

And you know which drivers you need to alert to get out of the way.

In veš katere voznike je potrebno opozoriti, da se umaknejo s poti.

And we wanted to do -- how can we best alert everyone?

In hoteli smo narediti -- kako lahko najbolje opozorimo vse?

How can these cars whisper, "You need to get out of the way?"

Kako lahko ti avti šepetajo "Moraš se umakniti s poti."?

Well, it depends on two things:

No, odvisno je od dveh stvari:

one, the ability of the car,

prva je zmožnost avta

and second the ability of the driver.

in druga je sposobnost voznika.

If one guy has a really great car,

Če ima nekdo res dober avto,

but they're on their phone or, you know, doing something,

ampak telefonira ali počne kaj drugega,

they're not probably in the best position

potem ni v najboljšem položaju,

to react in an emergency.

da bi reagiral v sili.

So we started a separate line of research

Zato smo začeli ločeno raziskavo

doing driver state modeling.

o modeliranju voznikovega stanja.

And now, using a series of three cameras,

In zdaj, z uporabo treh kamer

we can detect if a driver is looking forward,

zaznamo, če voznik gleda naprej,

looking away, looking down, on the phone,

stran, dol, na telefon

or having a cup of coffee.

ali sreba kavo.

We can predict the accident

Lahko predvidimo nesreče

and we can predict who, which cars,

in lahko predvidimo kdo in kateri avto,

are in the best position to move out of the way

sta na najboljšem položaju, da se umakneta

to calculate the safest route for everyone.

za najvarnejšo pot vseh.

Fundamentally, these technologies exist today.

Temeljno, te tehnologije danes obstajajo.

I think the biggest problem that we face

Mislim, da je največja težava s katero se soočamo

is our own willingness to share our data.

naša lastna volja za izmenjavo podatkov.

I think it's a very disconcerting notion,

Misli, da gre za zaskrbljujoče mišljenje,

this idea that our cars will be watching us,

ta misel, da nas bodo avti opazovali,

talking about us to other cars,

da bodo govorili o nas z drugimi avti,

that we'll be going down the road in a sea of gossip.

da se bomo vozili po cesti v morju govoric.

But I believe it can be done in a way that protects our privacy,

Ampak verjamem, da je lahko storjeno tako, da je naša privatnost zaščitena,

just like right now, when I look at your car from the outside,

tako kot zdaj, ko pogledam vaš avto od zunaj,

I don't really know about you.

v resnici ne vem nič o vas.

If I look at your license plate number,

Če pogledam vašo registracijo,

I don't really know who you are.

ne vem kdo ste.

I believe our cars can talk about us behind our backs.

Verjamem, da lahko naši avti govorijo o nas za našimi hrbti.

(Laughter)

(Smeh)

And I think it's going to be a great thing.

In mislim, da bo to dobro.

I want you to consider for a moment

Hočem, da razmislite za trenutek,

if you really don't want the distracted teenager behind you

če res nočete, da bi raztresena najstnica za vami

to know that you're braking,

vedela, da zavirate,

that you're coming to a dead stop.

da se popolnoma ustavljate.

By sharing our data willingly,

S tem, da delimo podatke,

we can do what's best for everyone.

storimo kar je najboljše za vse.

So let your car gossip about you.

Torej, dovolite, da vas vaš avto opravlja.

It's going to make the roads a lot safer.

S tem bodo ceste veliko varnejše.

Thank you.

Hvala.

(Applause)

(Aplavz)