self-driving cars are the future and I
don't think anyone questions that now
some people may not necessarily like the
idea of giving up control to a computer
but self-driving cars have the
theoretical potential to eliminate
traffic to decrease fuel consumption and
to transport more people more safely
than ever before Google way most
sponsored my trip out to their Chandler
Arizona Depot this past weekend where
they have hundreds of cars roaming the
streets autonomously and while the
incredible technology that we'll talk
about is better than it's ever been
before I still think that we are many
many years away from having cars capable
of truly driving themselves parked in
our garage why well let's find out this
video is sponsored by privacy car create
virtual credit cards that protect your
money sign up today with the link below
the Society of Automotive Engineers
partnered with the National Highway
Transportation and Safety Administration
that's a mouthful have outlined five
varying levels of autonomy and it's
important to note the distinction
between each level now the first three
levels they still require a human driver
with only the last two being high
automation or full self-driving now many
vehicles today ship with level one
driver assist hardware as a standard
feature like radar cruise control now
more advanced systems like Tesla
autopilot and Cadillacs super cruise are
generally considered by experts as a
high level to automation if you're
interested in what the cars you can buy
today are truly capable of be sure to
check out my video I did a few months
ago on Tesla's autopilot capabilities
they still require the constant
attention of the driver and because of
that they're not level three and they
can only operate dynamic driving tasks
in very specific scenarios such as on
highways that said when they are in
their controlled limited environments
they generally manage all driving tests
such as speed steering traffic awareness
etc now there's a big jump in between
level 3 and level 4 systems level 4
systems are being tested internally by
over a dozen companies but nobody gets
recognition for full self-driving more
so than way moe and for good reason what
started as a Google X project is now its
own independent
company under alphabet in nearly ten
years later wham-o has no problem
bringing press like me to check out
their cars they're even actually
operating a closed beta taxi service to
the citizens of Chandler Arizona pretty
cool their cars use four types of
sensors built in-house and retrofitted
onto their fleet of more than six
hundred sexy white Chrysler Pacific and
minivans okay so maybe not that sexy but
they are packed with amazing tech most
obvious is the funky-looking ball cap or
took for my Canadian viewers with eight
optical cameras that can see 360 degrees
around the vehicle now they also use
front and rear radar sensors to detect
objects in the rain the snow and the fog
or vision is basically reduced most
notably however and this is something
that is exempt from basically all
vehicles being sold today but is present
on Google Way MO and many other really
advanced off driving projects is lidar
now think radar but using lasers unlike
Tesla lame-o believes that lidar is
absolutely necessary for safe level four
and up autonomy lame-o says that it's
light our system which it's built
in-house can reliably map visual depth
which optical cameras and even our
eyeballs kind of struggle with and it's
accurate enough to see a helmet more
than two football fields away in
distance pretty crazy but regardless of
methodology all of these companies are
trying to create what's called sensor
fusion or the idea that combining data
from multiple disparate sources
increases the computer certainty of the
object's existence basically three
different types of sensors confirming
the existence of an obstacle is much
better than one confirming the existence
of an obstacle
especially since each sensor technology
has different strengths and weaknesses
but you see gathering data it's kind of
the easy part actually interpreting it
is where it gets really hard look us
monkey brains humans we're not really
very great drivers we're becoming
increasingly distracted by our
smartphones and we're even horrible at
basic principles of driving like
properly zipper merging keeping right on
the highway etc a lot of things that
could reduce traffic but for as bad as
we are at driving we really don't get in
accidents
often at least not on average because
there are more than six million car
accidents that occur in the United
States alone every year however that's
only an average of about 519 accidents
per 100 million miles driven which means
99.999 five three nines five percent of
the time that we're driving we don't
crash self-driving cars need to be at
minimum as reliable as the average human
driver before mass adoption really makes
sense
Elon Musk said 99% accuracy is easy
99.9% accuracy is extremely difficult
and each added nine after the decimal
point is orders of magnitude more tricky
if not impossible and we need at least
three point five nines now wham-o may
have very well surpassed that threshold
because after 10 million miles driven
their system has never caused an
accident the average driver would have
crashed more than fifty three times in
that same distance now way most still
puts employees in the cockpit for safety
reasons but I went on a 15-minute drive
without him intervening even once in
fact I was the one that pushed the start
ride button inside the car so he did
literally nothing the car handled the
stop signs traffic lights cross traffic
even parking lots without issue it
didn't drive as smoothly as a human
driver acceleration lane changes and
especially turns were at times just a
tiny bit twitchy robotic well really did
it was kind of just like a really really
safe teenager who never gets in
accidents not a bad thing but not quite
perfect now that said I never felt
unsafe and I would absolutely fall
asleep in that car driverless without
any concern well so then what's the
problem sounds like we're ready to
rumble well you see wham-o is level four
but only level four in areas that have
been specifically hand mapped that's
right
Google drives around and automatically
collects a bunch of data but then they
have to use human labor to actually
build out their mapping structure and I
probably shouldn't make any assumptions
but given the fact that they're
operating to the public in a very small
suburb of Phoenix it makes me think that
this requires a fair amount of human
capital
to build out and that's after more than
eight years since Google initially
revealed their low level for
self-driving vehicles I'm seeing this
all because despite what certain CEOs
proclaim I don't foresee mass level five
self-driving and hitting the roads by
the end of this year in fact I could be
wrong but I still estimate that it's
more than a decade away and when I asked
industry experts what they thought they
said my estimation was probably still a
little bit too optimistic I often people
see proclaimed that the tech is ready
we're just waiting on legislative
approval the government is slow ready
sorry pal but that's not entirely
accurate now law in the United States is
typically reactive however some states
such as mine Utah have already
proactively passed legislation that
gives level four and level five vehicles
full permissive use even for passengers
without a driver's license so it's it's
not legislation that's the easy part
it's the tech we're still not there yet
now Google way mode is incredible you
can't help but be amazed as you sit in
the backseat and watch the car drive
itself fully autonomously it truly feels
like you're in the future it will be a
game-changer it will be a black swan it
will change the world
someday but not today because today is
all about privacy calm the best way to
secure your purchase is online but
that's not a segue I don't know what it
is privacy allows you to protect
yourself online by creating virtual
disposable card numbers you can create
burner cards that you can set spending
limits on and you can even create
merchants specific card numbers for
example this is one that I have with
Netflix it has an $11 per month limit
and can only be charged by Netflix
nobody else even with the full credit
card number
privacy lets you use any billing address
you want and permits opening and closing
as many card numbers as you please
privacy makes money from merchants so
you never have to spend a penny try
privacy calm out today with the link
below and get five dollars free when you
sign up well folks that's all for me
thank you so much for watching
do you agree disagree let me know in the
comments down below but most importantly
and as always stay snazzy see you later
folks


0 Commentaires