I'm here to talk a little bit about one
of my absolute favorite topics to dork
out on, which is vehicle automation, and
before I dive in I'll just say that this
is emphatically a group effort and one
in which we have a really great team of
researchers at DOE and the National Labs and
in the broader community working on this.
So, I'll talk to a few things that we've
found so far, but I don't any point want
to make it sound like this is this is
just my work or I've been working on
this alone. So let's dive in. So the first
point I would like to make is that
transportation is great, transportation
is really amazing, we actually heard from
Jose that we need transportation even
for a wind industry so it turns up
everywhere. We might be more familiar
with some personal some personal uses
that we can now take planes to get
really anywhere in the entire world in a
day or two, we can rely on a global
freight network to make our goods
cheaper, we can use our highway system to
see beautiful places all over the US -
this is from Big Sur a place that's
particularly meaningful to me, and then
transportation supports the modern
cities. So this is New York City on the
bottom with trains with transit with
transportation options such as now we're
seeing like around DC Capital Bikeshare
and really giving us a wealth of
different ways to get around, move people,
and things throughout the country. So we
really have an extraordinary amount of
great transportation service, but then we
would be remiss if we pretended that our
transportation system doesn't still in
the United States at least largely
revolve around the automobile up at the
in the middle of that. And that's a
big part of the sort of corollary in the
second half of this dot, dot, dot sentence, which
is that transportation also creates
significant problems. So we know that
there's huge safety risks to
transportation, about 30,000 people in
the United States die every year in in
road crashes, congestion is a huge
problem nationwide -DC I think took the
trophy again this year, the last year
analyzed for the worst congestion the
most time lost due to congestion, we see
over 160 billion dollars of costs loss
to due to congestion from wasted fuel
and time sitting in traffic. There's
enormous equity issues raised by
transportation where we provide
transportation services that maybe don't
give equal access to transportation
services for everybody and then of
course the issues of pollution and
contributing to to man-made climate
change, transportations responsible about
a third of greenhouse gas emissions the
United States as well as quite a few
criteria pollutants maybe this is an
extreme
example but we do see this more broadly
system-wide. And then we have a huge
number of space issues that we have to
work through with transportation. This is
maybe a tongue-in-cheek image of a of a
parking space and how many more spots
you can take up with bicycles if you use
those a transport method rather than
cars and then on the right a
demonstration of the space occupied by a
group of polluted 50 people and their
different transportation modes, whether
they're using single akhiyan vehicles
bicycles or or or buses to get around. So
this is just to say that we have there's
a cost side of the ledger as well when
we're talking about transportation. So we
do have one technology that's been
rolled out as may be the solution to all
of society's ills in in transportation
and that's vehicle automation and some
people definitely think this is going to
be big and impactful so I'll just read a
couple quick quotes. So, "It'll be an order
of magnitude safer than a person. In fact,
in the distant future, I think people may
outlaw driving cars, because it's too
dangerous. You can't have a person
driving a two-ton death machine," and
that's from from Elon Musk of Tesla. From
Google, "A fully self -driving car has the
potential to have a huge impact. Mobility
should be open to the millions around
the world who don't have the privilege
of holding a driver's license." So really
focusing on the equity impact so the
benefits the automation could provide to
people who can't or choose not to drive
their own vehicle. And then lastly from
from the Department Transportation,
"Automated vehicle safety technologies
signal the next revolution and roadway
safety." So this is a part of DOT's
release of their first ever automated
vehicle policy that I'll talk a little
bit more about later, really signaling
that the the government believes that
this is a technology that's set to have
very profound, very profound impact. So
I'm going to focus today, almost
entirely on the energy impacts. That's
not to discount the safety equity
congestion all of these other impacts
which I think you're going to be
profoundly important but only have 15
minutes, which is a ludicrously short
amount of time to talk even about one
set of possible impacts and that's been
my area of research. So that brings us to
the central question, will automated
vehicles save energy and reduce
emissions or could they backfire? And
before I dive into that let me talk very
briefly about the technology and how we
got where we are. So a quick history. It's
really emerged almost out of nowhere on
the scale of how long these technologies
usually take to develop the modern
automated vehicle has its roots
series of challenges by DARPA - the
Defense Advanced Research Projects
Agency, from 2005 to 2007. First in the
desert and then in the urban environment
where teams of researchers from
universities and companies competed to
build vehicles that when started would
complete a course without any
intervention from human driver operating
completely unmanned. The first one in 2005,
0 vehicles completed the
test and then two years later in an
urban environment in 2007, six teams
completed it -the one shown on the left
and then DARPA declared victory and the
researchers behind those topics went off
into the world and largely went enjoying
the companies that we're now seeing
making a huge impact in the automated
vehicle world. The next huge landmark
that I want to call out was an
announcement by Google in 2012 that they
had already by that point driven 300,000
miles of self-automated driving without any
incidents at that point since their
program started in 2010 they were using
at that point modified aftermarket
vehicles they took hybrid vehicles and
rigged them up with a bunch of sensors
the lidar on top there a bunch of
cameras and a bunch of computing power
on board coupled with cloud access and
were able to operate those vehicles in a
wider and a wide and growing variety of
topics and they continued developing
that technology to spun off a company to
continue to develop automated vehicles.
And then the most recent milestone was a
2016 release of the first-ever
federally automated vehicle policy that
lays out the department of transportations
vision for how federal and state policy
will interact and how they expect to use
their policy making abilities going
forward. So not all these automated
vehicles are the same, they operate over
a wide range of different levels of
automation. I'm aware this is a total eye
chart, I start with it because there is a
little disclaimer here that says I'm
only allowed to use this if it's
reproduced as is so here it is as is now
we'll zoom in because this is maybe a
little bit more useful way to look at it.
We can differentiate the levels of
automation from zero all the way up to
5. Zero is what you call no automation, it's
actually hard to buy a car now that has no
automated vehicle features whatsoever
because level one automation driver
assistance is things like automatic
stopping in case of identifying
something when you're backing up behind
the vehicle or something as simple as
cruise control that can just maintain
speed on the highway that's actually a
kind of a vehicle automation system. Then
you start to get more into these
combined systems, so level
two it still you expect the operator of
the vehicle the driver to be paying full
attention but you would expect multiple
systems to be assisting the driver in
that case so this could be combining
lane-keeping with adaptive cruise
control them keep you the same speed and
in your lane but you as a driver would
still have to have your hands on the
wheel and be fully attentive. There's a
big jump when you go from blue to green
where now it's the vehicle automation
system that's monitoring the driving
condition and operating the vehicle, but
in level three automation and this is
critical the driver is expected to be
able to retake control of operating the
vehicle at any time. So you would have to
be constantly aware so Tesla's vehicle
when it drives on on highway is in this
mode where it's expecting you is the
driver be able to retake control at any
time that's critical.So, when you jump to
level four high automation, this is no
longer assumed that now the system is
designed under the assumption that human
not need to be always available to take
control and this is when some of the
more exotic effects that we're going to
talk about start to take over and then
the last level is level 5, full
automation this is true end-to-end
automation where you could actually get
in your car, have that vehicle take you
to your destination without any sort of
input other than asking the vehicle to
take you where you want to go. So I
mostly want to talk about these just to
say that state that a lot of the effects
on energy really only start to show up
in these high levels of automation and
that's why I'm going to be focusing on.
We do expect these sorts of vehicles
will mostly be a little further out,
estimates for when they might become
available very, but we already see and
I'll talk about this in a moment auto
makers that are proposing the leasing
level for level five vehicles within the
next decade. So this is something that
could potentially become an issue pretty
soon, we don't know how it'll how it all
play out. So there's a ferocious race
going on right now where every major
automaker is developing automated
vehicle features, I haven't talked to a
single automaker that says we don't
think this is going to be a big deal.
I've pulled six sort of semi random ones
to try to give a sense of the span so the
top left is this is Tesla, Tesla has
already have some self driving features
on the highway and have plans to release
more via software update. So they say
they now will sell vehicles that have
all the hardware in their opinion to
provide full self-driving and it's just
a matter of a software update as you go...
that's pretty crazy stuff. Ford, shown
here, has said they'll release level four
automation vehicles for sale in 2021 -
that's the date that's coming up very,
very soon. Um and then Uber people have
probably seen in the news is already
testing self-driving vehicles in several
markets with a few well-publicized
incidents, but they're already
demonstrating these vehicles on the road
and they see this as a critical part of
their business model going forward. These
though on the top are still sort of
sedan models, they're automating
conventional sorts of vehicles, there are
some more exotic designs out there. The
bottom left is a Mercedes concept
vehicle they demonstrate at the Consumer
Electronics Show where it's essentially
a mobile living room you can see the
front seats can actually rotate back so
they don't even face the front of the
car anymore and you could be expected to
sort of do whatever you want as your
vehicle it gets you around. They're quite
large and I don't believe they put out
fuel economy numbers, but we wouldn't
necessarily expect this to be the most
efficient vehicle on the road. Maybe on
the other side of the spectrum, this is
the new Waymo, which is the spin-off of
Google and their self-driving car their
demonstration vehicle that they're
operating now in in select areas as a
test vehicle they've designed this to
not even have a steering wheel and a set
of pedals. So the idea is this would be a
small vehicle that would come pick you
up was part of a service fleet take you
where you want to go and drop you off
again, a very different sort of, very
different sort of business models in the
Mercedes vehicle. And then I won't talk
much about it today, but there's also a
ton of freight demonstration options out
there vehicles like this that would show
up and deliver your package state
straight door-to-door maybe getting
towards that that Amazon Instant that
they're obviously pushing for. I'm not
going to focus on this today but those
have enormous potentials to reshape the
energy landscape as well so I mainly
want to put these up here to emphasize
that there's a huge diversity of
expectations among automakers for what
sort of vehicle consumers are going to
want and what sort of a business model
this is going to serve into. Are they
going to be fleets of automated vehicle
picking people up? Or are they going to
be privately owned vehicles much like
the mode that we have today? So let's go
back to the central question -will
automated vehicle save us energy and
reduce emissions or could they backfire?
So the short answer is we don't really
know yet, we've looked at it pretty hard
and DOE is working on it. I'll go into
a little bit more detail because I know that's
kind of satisfying. So DOE has
recently started to expand its efforts
on transportation to go beyond the
individual vehicle and actually look at
the whole system. So looking at systems
effects like urban science, like advanced
fueling infrastructure, multimodal
interactions, and then on the top right
connectivity automation asking questions
such as what are the energy implications
of connectivity and automation, and
that's what all that's what go through
today. So, what are a few of those effects?
We can divide up the possible effects
that my colleagues and I have identified
by look into demand and efficiency
changes. So on the demand side, we expect
that we can have some changes to vehicle
miles traveled or vmd how much people
drive, as well as changes in mobility or
access to transportation services. On the
efficiency side, we expect to be
affecting the miles per gallon, the
efficiency that a vehicle can operate, as
well as changes in operation -the context
in which they're operating and then
not not shown here you also might expect
automated vehicles to be suitable for
electrification if the vehicles know
where to go to plug into charge, if they
can match the range of the vehicle to
the trip that's being acquired. So I
won't go through each of these, just for
lack of time that the reference down
here and then I'm also happy to talk to
any of these with people, I'll just talk
about a couple. So on the demand side, you
expect travel to get easier, you expect it
to get cheaper, and you don't necessarily
mind as much the time that you have in a
vehicle if you're able to use that time
productively, that could be checking your
email, that could be watching a movie, that
could be taking a nap. We expect that if
people don't mind spending time as a
vehicle as much, they might travel
further or potentially live further from
their destinations. You might also shift
more more miles into into vehicles from
other modes,
such as such as aircraft or
transit potentially. On the plus side, you
might reduce the amount of time you
spent hunting for parking. People spend an unbelievable amount of time,
especially in the city environment,
circling around looking for parking
spaces and you might you might be able
to reduce that. You might also be able to
enable significant improvements in
ride-sharing ...if that's a seamless
experience where you're able to jump in
and out of a vehicle without it without
it being too complicated we're already
seeing business models like that
starting to emerge. On the efficiency
side, you might be able to smooth out
your drive cycle so we know that a human
driver who drives sort of smooth
acceleration and deceleration is much
more efficient automated vehicles are
very good at that you might be able to
do platooning so this is where you'd
have two vehicles or more on the highway
especially operating very close to each
other, closer than it's safe to operate
up with a human driver where they're
usually connected and then working
together very quickly to stay the exact
same distance with that slipstreaming
effect you can improve the mileage.
Anybody who's ever ridden a bicycle
behind somebody else knows how much
easier it feels to do that when somebody
is is getting the air out of the way for
you. And then on the downside, on the
efficiency side, you might enable very
significantly faster travel safely
that's
great from a time-saving perspective, but
from an energy perspective the faster
you go the more the more energy use. So
we as a team have begun to quantify
these in 2014 release the paper trying
to quantify the 11 possible effects we
felt there's literature to do so I won't
again go through each of these but i
just want to point out that this is a
percentage this is a fractional change
so we're looking at increases of up to
fifty percent from some of these
individual effects and savings very
significant on the other side as well. So
when you add these up, and this is I
think the analysis result I'm still the
most teased about, we found that the
energy savings the petroleum admissions
reductions could be as high as ninety
percent depending on which affects
manifest themselves or you can end up
tripling your energy use and
transportation and energy submissions
and transportation, right? So that's
incredibly broad range so it's easy to
see why people but people tease me about
that a little bit, but I would argue that
that's actually an important starting
point to know how wide apart those
bookends are. So further analysis by my
colleagues has started to break that
down where you'd expect to see it. So in
conventional vehicles you might we might
expect a total US fuel use of about a
hundred billion gallons per year, partial
automation can add some uncertainty that
you could get some savings from some of
these partial automation effects but
it's really when you get into the full
automation with and without ride-sharing
that the full range starts to start to
impress itself upon you. So what this
emphasizes for us, is that it's really
going to depend on how these scenarios
and how these scenarios evolve. So I want
to break this down from a little bit
more of a storytelling perspective into
a sort of peril and a promise scenario since
that was in the title of the talk I
felt like I had to go to that somewhere.
So you can imagine a peril scenario
being something like faster travel and
productive use of time supports
sprawl, people are willing and happy to
live much further from their
destinations large vehicles with poor
fuel economy tend to dominate and we're
talking about individual ownership still
dominating and then lots of empty
vehicle miles maybe while its circling
around. So you can imagine maybe my day
starts, I have breakfast at home, I live a
hundred miles from from DC to the west
and a lovely little place in the hills, I
jump in my personally owned mobile
living room, take the you know hour-long
ride to work, going very fast on the
highway while I'm catching up on my
email and don't mind I can get all that
done before I even get to the office.
Then maybe I go out to happy hour after
work but its twenty bucks an hour to
park downtown,
so i'll just send my vehicle driving
around the block, waiting for me while I
while I'm in there and it'll come pick
me up and you know maybe there's no
rules against that and then it takes me
back home and so by maybe by the end of
the day my vehicle has put two or three
hundred miles on it it compared to you
know what would normally be a much lower
total today. On the flip side of that
would be the kind of promised scenario
where you mostly have shared use, you
mostly have city use, match the vehicle
to the trip so you can bring the right
size and range of vehicle to whatever
trip you need, you can electrify most of
those miles and then you have an urban
form that's adapted around automation. So
in this scenario, maybe I get up in the
morning, grab breakfast of the corner
cafe, when I'm ready to go to my first
meeting a little one or two seater
vehicle with 20 miles of electric range
picks me up and takes me to my meeting,
when I'm done with work maybe my
co-workers and I want to go out for a
drink so I maybe six of us are going so
a six-seater pulls up and takes us to
where we want to go so we're able to
socialize on the trip. So that's the kind
of energy promised scenario what I want
to emphasize about this is that both of
these scenarios are kind of awesome in
some ways both of these are providing a
energy transportation service that you
can imagine people being really excited
about so I'm not trying to say one of
these is better or worse the peril and
promise is really from an energy
perspective and something that we need
to think through in advance as this
system involves. So I would argue that
this is going to be more so and more so
not a technology question but a question
of how humans respond to these
technologies. So I've got a couple of
human questions I go through them very
quickly since I know we're already over
time and I hope I'm not keeping people
from next meetings you can have a sort
of promised an area of platooning road
trains where highways are dominated by
vehicles working together to save energy
but on the other side if you're at the
front or the back of one of these you're
not getting those energy saving benefits
so what's to stop me from writing an
algorithm that tries to get myself into
the middle of those well then you don't
have a Road train that's going to work
very well right so the broader question
is will we write selfish algorithms for
these vehicles will be like cooperative
algorithms to these vehicles... I don't
think we know the answer to that yet.
Will we have urban travel on demand
models or this sort of sprawl long
commute model? And this is really going
to depend on where people want to live
in an automated future. People choose
where they want to live that that that's
how we end up dictating our form so it's
really about what sort of future we're
looking for. We can have the right-sized
sort of protocol
right sides bringing the small vehicle
when you do smaller at the large vehicle
for a large trip versus the cell phone
mobile lounges and here we really are
asking what sort of vehicle two people
want what will people choose based on
how much they cost. There's a great
promise in what we call the marginal
pricing of travel, so this is if I don't
own a vehicle I don't have that sunk
costs have already having spent the
money on the upfront I'm actually just
spending the money per trip and that can
change how you perceive travel in the
amount that you travel, but if we're
going to very cheap travel as we said we
might expect to see a very significant
increase in vehicle miles travel. So this
gets us questions like, do I want to own
a car under these circumstances? Should I
travel by car or are there other modes?
Do i do I want to travel a different way
or do I not want to travel? And then the
last one, which I think really is a good
framing for how humans think about these
questions. We know we could design
an algorithm with an efficient button
that would say I'm going to take you the
most efficient route, I'm going to take
you the most efficient speed, i'm going
to start and stop smoothly or we could
design an algorithm in a fast button
that says get me there as quickly as
possible. Which of those do you think if
you have those you would press and then
which do you think we as a society will
press? And I don't have a question and
answer to any of these questions on
these are just I think the things that
are going to define what the the future
of energy looks like. So what's next?This
is being worked out right now the state
level. This is a map that Stanford
produces of states that have taken on
past or not yet passed a legislation
relating a self-driving cars state
policy isn't really built for them when
having to figure this out as we go along
this is being developed in live time,
federal policies also being worked out
at the same time, cities are having to
figure out what this will look like and
companies as I said are developing this
as a frantic pace. So we don't know the
answer to any of these questions yet,
what I would argue is that I believe
based on the evidence we've collected so
far that automated vehicles represent
one of the essential technologies that
we might be able to use to combat human
caused climate change we shouldn't take
it for granted though but in closing we
are I believe still in the driver's seat
and we have a choice as we go forward.
Thank you very much.
For more infomation >> Citroën DS3 1.6 e-HDi So Chic - Duration: 0:58. 
For more infomation >> Mercedes-Benz E-Klasse Estate 200 CDI Ambition Elegance Aut,Leer,Navi - Duration: 0:55. 
For more infomation >> Citroën C3 1.6 E-HDI DYNAMIQUE NAVIGATIE/ ECC/ TREKHAAK - Duration: 1:02. 
For more infomation >> Mercedes-Benz E-Klasse 230 Elegance Aut,Leer,Schuifdak*YOUNGTIMER* - Duration: 0:56. 
For more infomation >> Peugeot 308 SW 1.2 E-THP 130PK ALLURE - NWSTAAT - Duration: 1:01. 
For more infomation >> Peugeot 108 5DRS E-VTI ACCES - AIRCO - CD/AUDIO - NOIR - Duration: 1:04. 
For more infomation >> R.C. Sproul - Bárkinek hirdethetjük-e, hogy "Isten szeret téged?" - Duration: 3:34. 
For more infomation >> Existenzgründung mit e-Trikes – das ELFIT-Lizenzsystem in 60 Sekunden erklärt - Duration: 1:04. 
For more infomation >> Citroën C3 E-VTI TENDANCE - AIRCO - 14DKM - TOPDEAL - Duration: 0:54. 
For more infomation >> Peugeot 208 5DRS 1.2 E-VTI ALLURE-NAV-PANORAMA - Duration: 0:54. 
For more infomation >> Citroën C1 1.0 e-VTi 68PK airdream 5D Airscape AIRSCAPE FEEL - Duration: 0:54. 
For more infomation >> Citroën C4 Cactus 1.2 E-VTI FEEL - AUTOMAAT - AIRCO - Duration: 1:00. 
For more infomation >> Citroën C1 E-VTI 5DRS FEEL AIRCO/USB/BLEUTOOTH - Duration: 0:49. 
For more infomation >> "Lo que queda claro es que nos hemos convertido en un país de fosas": Denise Dresser sobre México - Duration: 8:36. 
For more infomation >> Fabrício Zava – Sem Lógica - INTERSECÇÕES (Áudio Oficial) - Duration: 2:26. 
For more infomation >> A arquitetura da desigualdade - Duration: 2:31. 
For more infomation >> Rochelle - You Got Something (PONKERS PARODIE) #GEINIG - Duration: 3:23. 
For more infomation >> Gravata de Vidro - O Nome do Vento [Live in Hotel Mercure] - Duration: 3:48. 
For more infomation >> Opel Corsa 1.0T 66KW/90PK 3D Online Edition - Duration: 1:00. 
For more infomation >> Opel Zafira Tourer 1.4T 140PK Automaat Innovation / Navi - Duration: 1:06. 
For more infomation >> Opel Astra 1.4 T 150pk 5-drs Innovation / Full Option - Duration: 1:02. 
For more infomation >> Dmx - Ain't No Sunshine - Duration: 5:29. 
For more infomation >> LE SUPER SAIYAN BLUE EST DEVENU INUTILE ? - Duration: 6:41. 
For more infomation >> 遊戲驢子評論Yooka Laylee(驢子觀點) Yooka Laylee (dunkview) 中文字幕cc - Duration: 6:30. 
For more infomation >> 6 TOP THINGS To Do In LONDON! 🇬🇧🙌🏻 | Travel Better with Holiday Extras! - Duration: 1:53. 
No comments:
Post a Comment