ADAS & Autonomous Vehicles Special Report - December 2024 - 22

TECHNOLOGY REPORT
eventually will replace humans,
McEwen said. " That may seem
farfetched, but within two years
Amazon will have more robots in
their warehouses than humans. So,
that inflection point is coming very
quickly, " he said. " What I would like
to see AI used as is a co-pilot, which
is simply an augmentation. Think
of it as increasing your intelligence
by 10 IQ points and increasing your
productivity by 25%. You're going to
see many more of those applications in
the short term across every industry. "
Gary Johansen, VP of engineering for
power systems at Cummins, discussed
his view on AI's capability to improve
product development. " I think if you look
across the entire value chain of what we
do in product development, one of the
best uses for this tool is alleviating the
engineer tedium. Relieving the engineer
of tedious tasks of report generation
or documentation is very helpful.
" At the same time, we've got to be
careful so we don't remove [engineers']
opportunity to do their own thinking, "
Johansen continued. " If you move
all the way up the value chain to the
product creation, product conception
and exploring the design space in
hardware in addition to software,
and maybe finding solutions that we
perhaps wouldn't have thought of. So,
one of the things we're trying to do
right now is make sure we're not just
focusing in one part of the value chain. "
SAE COMVEC executive panel on AI (from left): Alan Berger of abcg, Volvo CE's Ray Gallant, Steve
Greenfield of Automotive Ventures, Phil McEwen of Continental, and Cummins' Gary Johansen.
Cummins is " walking before running "
when it comes to AI implementation,
Johansen added. " We're trying to
sample and pilot in areas that give us a
strategic view about how these things
might help us all the way through value
creation. We're not to that point yet,
but I think in terms of saving money,
saving time and creating value, we're
certainly trending in that direction. "
Predicting trends
Volvo CE's Gallant provided
observations on AI's capacity to analyze
large swaths of data and spot future
trends. " AI is very important not only
for analyzing huge amounts of data and
spotting trends, but also for determining
how can we use those to predict future
trends. That's how we get ahead of what's
probably going to happen, and AI is very
good at building those kinds of models
in real time by using the latest data that
we couldn't possibly do on our own. "
AI also has the potential to remove
some of the biases engineers may
have, Gallant said. " We're all trained
practically as engineers in our
problem-solving routine to find the
root problem. Then you come up with
possible solutions, implement the
solution, test it, then go back and follow
up and tweak it where necessary. "
" We all have inherent biases, " Gallant
continued. " Whether that's things
that have worked well for us before
or things that we're familiar with. And
we ignore technologies that may be
right beside us, but we're not familiar
with. AI has got a great promise
of being able to look at all those
technologies and propose solutions
that may never have entered our sphere
of influence if you will, when we're
coming up with those solutions. "
Gallant also provided a caveat to
Cummins' Acumen
advanced computing
device uses machine
learning and advanced
analytics to enable
remote diagnostics,
over-the-air
programming and
digital service tools.
22 DECEMBER 2024
the SAE COMVEC audience. " The one
caution I want to point out with AI is that
it is only as good as the data going into
it, " he stressed. " We're seeing examples
of misinformation where AI is providing
flat out wrong information because the
data in was bad. And of course, if you let
AI scroll the internet for all the data on a
particular subject, you're going to get a
lot of bad data coming in and your result
is going to be similarly questionable. "
Matt Wolfe
ADAS & AUTONOMOUS VEHICLES SPECIAL REPORT
FROM TOP: SAE/MATT WOLFE; CUMMINS

ADAS & Autonomous Vehicles Special Report - December 2024

Table of Contents for the Digital Edition of ADAS & Autonomous Vehicles Special Report - December 2024