Automobility LA - S6

autonomy

Road to a self-driving future likely longer than once imagined

ive years ago, automotive and technology executives freely speculated
that autonomous vehicles would be on the road by 2025.
Today, executives are less likely to venture predictions about Level 5,
or fully self-driving, cars-not only because of the engineering and technology challenges but also because the industry has realized it has a long way
to go in building consumer trust in autonomous vehicles.
The J.D. Power Mobility Confidence Index for the third quarter of 2019
placed consumer confidence in self-driving vehicles at 36 on a 100-point
scale. Any number below 40 is deemed low.
As more work has been done and more tests conducted-with the occasional accident and resulting negative publicity-companies have backed
off promising fully autonomous driving in the near future.
"If you ask now when are we going to see Level 5 widely available, I'd say
probably never, or at least not in the near future," says Dr. Carsten Breitfeld, an
auto industry veteran and global CEO of Faraday Future, a California-based
electric-vehicle manufacturer. "There's a chance to make
fully autonomous vehicles work in very clearly defined
areas, in 95% of weather conditions.
"But what I'm looking at is more of a Level 3.8 in the
next five years."
Steve Surhigh agrees. He points out that Level 2 auSurhigh
tonomous features-including lane assist, adaptive
cruise control and braking and acceleration support (Think Tesla Autopilot)-are integrated into about 10% of vehicles today.
"But we're still expecting true Level 4 will be a ways off," says Surhigh, vice
president and general manager of cloud services for Harman International.

6 · AUTOMOBILITY LA · NOVEMBER 2019

"There are a number of factors that influence that, including how much has
to be invested to get to Level 4."
At Level 4, the car should be able to drive itself safely, even if the human
driver doesn't respond to a request to intervene.
Just as important as the technology is consumer interest in autonomous
vehicles, Surhigh says.
"A big factor is the willingness of riders to trust the vehicle," he says. "That's
where we're trying to help cross that chasm to get to Level 4 by helping to
change the experience of how drivers engage with the vehicle as they move
through progressive levels of autonomy."
Level 4 autonomy requires handoffs between driver and vehicle. Surhigh
says Harman and others "want to do that in a way where drivers are ready
to take back control when necessary, so that it's seamless and safe-not
clunky-and everything that we all want it to be."
Harman has an entire design team focused on the user experience in a
more autonomous-driving situation, figuring out "a balance between being
intrusive with information and alerts and what's needed to be safe," he says.
"Trust is a key issue," says Tao Liang, director of machine learning at NIO, a China-based manufacturer of
electric and autonomous vehicles. "Different means can
be used to build a bond and trust between the user and
the vehicle itself, including artificial intelligence.
"We're being proactive in terms of setting expectaLiang
tions in terms of how the vehicle and the driver would
interact during a trip. There are ways to build that consumer trust, and Nomi
is the face of that for us."

Automobility LA

Table of Contents for the Digital Edition of Automobility LA