WEBVTT

00:00:00.080 --> 00:00:06.960
i don't think anything says we're living in the future quite like the

00:00:04.000 --> 00:00:11.840
proliferation of self-driving cars now granted they're not exactly all over the

00:00:09.040 --> 00:00:15.759
roads yet but given the huge investments that automakers and tech companies are

00:00:13.920 --> 00:00:22.800
making in them it seems like it's a matter of when rather than if

00:00:19.439 --> 00:00:25.039
but how exactly do they work i mean most

00:00:22.800 --> 00:00:28.720
of our trains aren't even self-driving and those things are quite literally on

00:00:27.519 --> 00:00:33.680
rails well as you'd expect self-driving cars

00:00:31.359 --> 00:00:38.320
are jam-packed with tons of equipment to help them both see and understand their

00:00:36.480 --> 00:00:43.280
environment some of it is rather familiar such as a data connection for

00:00:40.559 --> 00:00:48.399
traffic info and a gps transponder to allow the car to know where it is but

00:00:46.000 --> 00:00:53.120
while gps is fine for providing driving directions in a human-controlled car it

00:00:50.960 --> 00:00:58.719
has a margin of error of up to several meters and besides to control a car with

00:00:56.000 --> 00:01:04.320
such a slow and error prone system would result in more than a few fender benders

00:01:01.359 --> 00:01:08.799
so to start with self-driving cars also use freaking laser beams to build a map

00:01:07.200 --> 00:01:13.119
of what's around them lidar which you can learn more about up

00:01:10.400 --> 00:01:17.840
here can be built into units that can spin

00:01:14.280 --> 00:01:20.320
360 degrees shooting invisible lasers in

00:01:17.840 --> 00:01:25.360
all directions then measuring how long it takes for each beam to hit an object

00:01:23.040 --> 00:01:30.080
and bounce back this allows the car to judge not only the distances but even

00:01:27.759 --> 00:01:34.479
the shapes of the objects immediately around it giving the car even more

00:01:32.159 --> 00:01:38.479
information is traditional radar to help it gauge speed gyroscopes and

00:01:36.400 --> 00:01:42.320
accelerometers to provide more movement data than would be possible with a

00:01:39.920 --> 00:01:46.240
traditional speedometer and high resolution cameras now you might think

00:01:44.960 --> 00:01:50.720
well we've already got frickin laser beams

00:01:48.479 --> 00:01:55.040
what are the cameras for well although the car's other systems give it a pretty

00:01:53.119 --> 00:02:00.479
good idea of what's in the general vicinity the cameras really help provide

00:01:57.920 --> 00:02:04.479
a complete picture as they can see in color this helps your car distinguish

00:02:02.479 --> 00:02:10.080
let's say a caution sign from a construction sign so you put all of this

00:02:06.960 --> 00:02:12.480
together and the car has a coherent 3d

00:02:10.080 --> 00:02:17.680
map that provides the data that it needs to make decisions for example lidar and

00:02:15.760 --> 00:02:20.879
video can determine whether that thing up ahead is telling the car to stop or

00:02:20.000 --> 00:02:27.520
yield whether the vehicle in front is a small sports car or a large truck so that it

00:02:24.879 --> 00:02:32.400
can decide whether to pass or whether that two-wheeled contraption is a

00:02:29.520 --> 00:02:35.440
motorcycle or a bicycle so the car can get around the cyclist and give her a

00:02:34.000 --> 00:02:39.599
bit more space also many self-driving cars have

00:02:37.440 --> 00:02:43.680
ultrasonic sensors in the wheels so the car will know how close it is to the

00:02:41.280 --> 00:02:49.200
curb and other vehicles while parking and even microphones so that they can

00:02:46.319 --> 00:02:53.200
hear a police or ems siren to get out of the way in a timely manner i mean that's

00:02:51.760 --> 00:02:59.920
going to be one improvement on the roads around here once self-driving cars are ubiquitous people can be so rude they're

00:02:57.360 --> 00:03:04.239
going to an emergency anyway that's how all the data gets collected

00:03:02.319 --> 00:03:08.319
but what about the processing an experienced driver can take in

00:03:06.239 --> 00:03:12.000
information from the environment filter out everything that's not important and

00:03:10.319 --> 00:03:15.120
make all the kinds of decisions we talked about before in a fraction of a

00:03:14.159 --> 00:03:19.760
second so of course then you need a lot of

00:03:17.440 --> 00:03:25.040
computing power for a self-driving car to pull off the same thing this is made

00:03:22.000 --> 00:03:27.680
possible by running many processors in

00:03:25.040 --> 00:03:31.519
parallel to crunch these numbers it's actually quite similar to how desktop

00:03:30.000 --> 00:03:36.159
gpus work these types of processors also happen to

00:03:33.760 --> 00:03:41.519
be more trainable with machine learning so for example you can teach a car what

00:03:38.799 --> 00:03:46.080
a pedestrian looks like by showing it a large data set with lots of photos of

00:03:43.599 --> 00:03:50.799
people crossing the road but what if a self-driving car didn't only have to

00:03:48.000 --> 00:03:54.879
rely on its own sensors and processors well another goal is to have these cars

00:03:52.720 --> 00:03:59.599
communicate with each other while they're on the road so they can

00:03:56.879 --> 00:04:04.720
actually tell other cars what they're doing and why enhancing safety and

00:04:02.959 --> 00:04:09.200
taking some of the guesswork away from the individual vehicles and if you think

00:04:07.040 --> 00:04:13.120
about it having many cars working together like this in unison could help

00:04:11.200 --> 00:04:17.040
with lots of other issues too like easing congestion by coordinating

00:04:15.040 --> 00:04:20.560
movement so that traffic can keep flowing smoothly

00:04:18.639 --> 00:04:25.680
also in the future especially once faster 5g connections become more common

00:04:23.280 --> 00:04:30.000
we could even see smart infrastructure that communicates with the cars i mean

00:04:28.080 --> 00:04:33.919
think about a parking garage that could tell a car that it's too full to

00:04:31.520 --> 00:04:37.759
accommodate it or transponders in construction zones that could tell them

00:04:35.520 --> 00:04:41.600
to slow down and prepare for narrower lanes of course we do still have a long

00:04:40.000 --> 00:04:45.759
way to go before this kind of tech is commonplace but given how many

00:04:43.919 --> 00:04:50.479
semi-autonomous cars are available for purchase today and how much money is

00:04:48.720 --> 00:04:54.880
being poured into this whole endeavor along with the ever-rising processing

00:04:52.560 --> 00:04:58.160
and data speeds we continue to enjoy the hope

00:04:55.759 --> 00:05:03.199
is that in the near future our car trips will be much easier and safer since well

00:05:01.040 --> 00:05:07.680
over 90 percent of car accidents are attributable to human error and speaking

00:05:05.440 --> 00:05:12.720
of self-driving cars today's episode was brought to you by ibm spectrum storage

00:05:10.400 --> 00:05:19.039
did you know that a self-driving car can generate up to 15 terabytes of data

00:05:16.080 --> 00:05:24.400
every hour and lots of this is actually not even used while they're driving so

00:05:22.000 --> 00:05:29.840
it's really common in the ai industry to have to ingest just tons of this raw

00:05:26.960 --> 00:05:35.039
data and transform it after the fact while the car's not even driving into

00:05:32.240 --> 00:05:38.560
intelligence and ibm is using spectrum storage to help automakers manage all

00:05:37.520 --> 00:05:42.400
this data so it's optimized for ai and machine

00:05:40.639 --> 00:05:47.360
learning with industry-leading GPU accelerated servers and ibm spectrum

00:05:44.880 --> 00:05:51.440
scale software-defined storage combined with groundbreaking performance and

00:05:49.120 --> 00:05:55.039
simplified deployment so get the fastlane to insights into

00:05:53.039 --> 00:05:59.919
whatever field you're in by learning more about data storage for ai and ibm

00:05:57.919 --> 00:06:04.000
storage solutions for autonomous driving at the link below so thanks for watching

00:06:02.160 --> 00:06:07.680
guys like dislike check out our other videos leave a comment with video

00:06:05.520 --> 00:06:11.919
suggestions and don't forget to subscribe and follow

00:06:10.000 --> 00:06:16.160
not this one i mean this one too but but but also Techquickie that's a button

00:06:13.600 --> 00:06:16.160
below