The
following is an opinion editorial provided by Doug Davis, senior vice
president and general manager of the Automated Driving Group (ADG) at
Intel Corporation:
For
more than 30 years, I have climbed out of bed to go work at Intel. But
never in those 30 years have I been more excited to do so than I am
right now leading Intel's autonomous vehicle team.
Don't
get me wrong - I've had an absolutely wonderful career and have worked
on some programs that have had a significant impact on the world around
us. But the chance to solve one of the most complex technology
challenges of our time, the opportunity to help the auto industry
reinvent transportation, the potential to save a million lives every
year - those things are unlike anything I've done before. They're the
reason I postponed my retirement.
I
have unwavering confidence that Intel will succeed in autonomous
driving. We have an astounding breadth and depth of experience and the
world's finest technology toolkit to apply to this challenge. We have
tapped resources from across the company and have added experienced
talent from the automotive industry. Our teams are operating in high
gear and will deliver the necessary technology breakthroughs. Here's
why:
First: We are already demonstrating great progress
Intel
technology is in hundreds of autonomous test vehicles on the roads
today. Not all carmakers are talking about who is powering the brains in
their test cars lest they give away secrets. But the fact is many of
them are using Intel. I encourage you to open their trunks, boots or
hatchbacks and see which tech company they are relying on the most to
provide the brains for their development vehicles.
Today we are also showing off one of the first of approximately 40 highly automated driving (HAD) cars promised from BMW, Intel and Mobileye this year. It was less than a year ago that the three companies announced
plans to bring highly and fully autonomous vehicles into series
production by 2021 through the development of a common platform. We are
now successfully demonstrating that platform and are preparing plans to
bring it to market for other OEMs and tier-one suppliers to accelerate
their programs. Stay tuned for more on that front.
Second: We are prepared for the data challenge
The
single most important factor in autonomous driving is data - how best
to process it, manage it, move it, store it, share it and learn from it.
From PCs to data centers and everywhere in between, no company's
silicon has analyzed, computed and moved more data than Intel's. As we
move down the road toward autonomous cars, the data challenge will
become much more complex and require new ways to work with data inside
the vehicle, throughout the network and across the cloud.
To
ensure we have the absolute right strategy to handle that data
challenge, we've installed the first of several planned data centers
dedicated to autonomous driving. These unique labs will be used for
algorithm development and training, as well as for understanding the
special infrastructure needs for autonomous driving data movement and
storage. Researchers will continually feed information from Intel's test
cars into these data centers to train neural networks and improve
machine learning algorithms. And, we're busy building similar labs with
customers and partners.
Which
brings me to artificial intelligence (AI). Mastering AI both inside the
car and in the data center will be essential to the autonomous driving
data challenge. Here it's important to remember that autonomous driving
isn't a game. When cars are thinking and acting without human
intervention, they must be able to do so in a safe and trustworthy way.
The artificial intelligence needed to make this happen isn't just
computer vision - think voice, decision-making, personalization and
preferences. Each of those AI workloads needs a different set of
algorithms and likely different kind of processing for optimum
performance. If all we needed was a supercomputer to handle the
autonomous driving data challenge, our work would be done.
Third: We've built big industries before and will do it again
I've
said this before: Autonomous driving will accelerate when the industry
comes together to align on common platforms and technologies. That
enables developers to go quickly and in volume while still
differentiating their solutions in software.
How
do I know this? Let's look at the PC and server industries as examples:
Before 1980, the computer industry was highly proprietary, serving
primarily researchers, big companies and hobbyists. This changed a few
years later when IBM built the first personal computer using mostly
off-the-shelf parts and an outsourced operating system. These early PCs
were the base for a standardized approach to computer design that led to
faster evolution of technology and 150x growth in two decades.
The
data center is similar. After mainframes came microcomputers that
enabled some flexibility and cost-effectiveness. But it was the
adaptation of PC technology that allowed data centers to evolve quickly
to handle the data that came with the internet. IDC reported there were
1.8 zettabytes of data generated in 2011, and the agency estimates we
will generate over 40 zettabytes by 2020. Only through standardized
solutions have we been able to grow the industry fast enough to keep up.
Which
brings me back to autonomous cars. There are plenty of naysayers who
think Intel's experience in enabling rapid scaling cannot be replicated
in autonomous driving. At the same time, there are plenty in the
automotive industry who don't understand how open collaboration can
enable differentiation and innovation. I understand the skepticism, but
from years of experience, I know that technology solves problems best
when it's organized around common platforms and predictable interfaces.
Without a doubt, that's the fastest way forward on our autonomous
journey.
As
a society, I don't think we can afford to continue down a proprietary
path. The cost in time, money and human lives is too great. The faster
we can deliver autonomous driving technology and take humans out of the
driver's seat, the faster we can save lives. It's that simple - and that
important. And I am confident Intel will not only succeed in helping
our partners put self-driving cars on the roads, we will do so in the
fastest, smartest way possible.