Shift Magazine - February 2020 - 23

"

It's not about
replacing a human
being; it's about
supplementing
and augmenting
their activities.

"

Dennis Clark, managing director of
strategic venture partnerships at
Honda Innovations

UVeye's cameras and scanners learn almost
instantly and can find anomalies right away.

Monolith AI applies deep learning to
product design.

Dennis Clark, managing director of
strategic venture partnerships at Honda
Innovations, said Honda has designated
a few factories around the world as test
sites for this kind of technology, and
that it is working with two startups that
have an AI focus. The first is Monolith
AI, which applies deep learning and AI
to the design of parts, specifically CAD
modeling. The second is UVeye, which
applies vision machine learning to vehicle
inspections.
"The idea for both of these technologies and the reason why we like them
so much is they augment and help our
workers in these environments to create
better products," Clark said. "It's not
about replacing a human being; it's about
supplementing and augmenting their
activities, and that's really where we see,
as a company, the value of AI. It almost
gives us superpowers."
Since late 2019, Honda has been
working on multiple, unnamed projects
with Monolith AI. Monolith AI uses
old Honda model design data to assist

Honda engineers who are updating
those models or working on new ones.
Honda is not giving out concrete details
about these projects, but Clark said
even though the partnership is fairly
new, "we have seen some early positive results."
Joël Henry, Monolith's senior AI engineer, would also not give specific details
about how his company works with
Honda to shorten design cycle time. He
did say, as a general example, that perhaps an engineer who would have had
to spend eight hours checking a design
for technical requirements in the past
can now let the AI help with this work,
automating the task thanks to machine
learning algorithms. "This helps because
[the engineer] will have more free time to
think of innovation rather than checking
things," he said.
UVeye's technology is also able to
check for things - in this case defects
or anomalies - as cars move down the
production line. UVeye uses multiangle
scanning technology to quickly build

digital models of a vehicle or a part
and then makes sure that there are no
problems with those vehicles and parts
as they are produced. The inspection
system can be placed anywhere on the
production line.
UVeye COO Ohad Hever said the company is working with a handful of OEMs
that have been publicly named - including Volvo, Toyota, Daimler and Skoda -
on top of "several discussions" that are
ongoing. The Honda partnership is one of
UVeye's biggest, Hever said.
"We can detect anomalies in the early
stages, so they don't need to find it in the
final stages and return the vehicle back,"
he said. "This is very helpful to them."
Hever said UVeye sets up its cameras
and scanners based on customer needs,
and since the system can learn almost
instantly, it can start finding anomalies
right away.
"All of the car manufacturers want
to introduce new technology into their
assembly lines," he said. "I think it's wonderful. For us, the more the merrier." n
february 2020 * shift

23
Shift Magazine - February 2020

Table of Contents for the Digital Edition of Shift Magazine - February 2020
