Wells Vehicle Electronics' Tom Schaefer Talks Manufacturing, Automation, + More

Introduction

Adam: This WiLL Cast is with Tom
Schaefer of Wells Vehicle Electronics. Becca, what'd you think?

Becca: He is such an interesting guy.
They don't make them like that anymore. He's been at Wells since — when did he
say? The mid-80s, I think. 1983–85, something crazy. And he had a really
interesting life before that. Very much self-taught, learned on the job. Had a
little bit of schooling, but for the most part figured things out as he went.
One of the things I kept thinking about was, man, this is exactly why we do the
WiLL Cast. It's as many people like this that we can get on and capture their
knowledge and their way of doing things, so over time people can see what's
capable and what you can learn. The knowledge that generation has — it's a
great thing to get on video.

Adam: He puts things in a historical
context, which I think is really cool. Talking about trends when he first got
started, how a lot of the changes in automotive electronics were driven from
government policy, then getting into consumer electronics and how that was
related to automotive electronics. The offshoring that happened the last couple
decades and now the onshoring that's happening. To be able to see that holistic
view of the American economy over many decades is fascinating.

Becca: Did you feel like there are a
lot of parallels with them and us?

Adam: Well, so much of the auto
industry and the lighting industry — we work with the auto industry a little
bit with some of the electric vehicle charging. They're a very high percentage
auto, but they have multiple channels to market, similar to us. They have more
of a retail channel, an indirect channel. With somebody like Wells, they kind
of show us what is capable with some of the manufacturing and automation. And
they're so open to having people through. Clearly we're not competitors — we're
potential customers. But it shows us what's capable in our industry.

Adam: This WiLL Cast is brought to you
by the Ledge Games, Becca's favorite time of year. Saturday, September 24th.
The last Saturday of the month. Events kick off right around 8 a.m. — axes are
flying, kegs are being tossed, hammers are being thrown. Visit LedgeGames.com
for more information. All the money made goes to support local builders, trades
people, education, nonprofits, and anything related to the core of Wisconsin
manufacturing here in Fond du Lac. Close to $45,000 has been raised in over six
years. It's entirely volunteer-run. It's a 501(c)(3). But enjoy this WiLL Cast
with Tom Schaefer from Wells Vehicle Electronics.

Tom's Background: DeVry, Motorola & Early Computing

Adam: So Tom, what's your background?
What'd you study in school? What's your career looked like?

Tom: School is interesting for me. I
wouldn't recommend this — I have a high school education. I went to DeVry
University for about a year and a half. We kind of ran out of money, so I
didn't get an associate's or a bachelor's degree. But I had five trimesters and
they gave you a diploma for that. But I could program computers in 1983, so I
got hired by Motorola's semiconductor research and development labs — not
probably because I knew enough about electronics, but because I could program
computers.

Adam: What did programming a computer
look like in the early '80s?

Tom: Punch cards. Actually, I did
punch cards in high school. My senior year in high school, I was a shop hound —
auto shop, electronic shop. Instead of taking calculus, I took Computer Basics
1 and 2, and that has served me for most of my career. From there you could
pick up microprocessors. The command sets were really simple. You didn't have a
big CRT in a lot of cases — you had essentially your calculator, single line,
20 characters to look at, green and black. You wrote everything down on a piece
of paper.

Tom: You get to Motorola — first
shift, that building was like parking at the Brewer game every day. There were
like 10,000 people there. They had computers and equipment that could be hooked
up to computers everywhere, and nothing was connected.

Adam: Was Motorola inventing the
technology and the chips going into computing products, or were you also
designing finished goods?

Tom: I think Motorola at the time was
supplying microprocessors to Hewlett-Packard. So Hewlett-Packard and Motorola
had this thing together. Their labs were all full of Hewlett-Packard gear. We
were working in areas that were kind of working on the first cell phone
technology. The PhDs and master's degree guys in my department were working on
all of that type of stuff. I had no clue what any of it was.

Adam: Was that the antenna side, or
were they designing circuit boards?

Tom: They were making the actual
semiconductor. It wasn't even silicon — it was a crystal called gallium
arsenide at the time. These little amplifiers and oscillators and mixers that
were going to create the high-frequency signal that you modulate the voice on
top of and transmit across the air. It was a really crazy environment.

1975: The Watershed Year for Automotive Electronics

Adam: What did automotive electronics
look like in the '50s, '60s, '70s? When did things truly start to switch from
electrical to electronics?

Tom: 1975 is kind of a
line-in-the-sand year. Pre-1975, cars from 1900 to 1975 — you essentially had a
switch, which was literally two pieces of metal that opened and closed. They
connected 12 volts to an ignition coil and then interrupted that, and when you interrupted
the current through that ignition coil, it produced a brief 30,000-volt spark
that would jump the spark gap. It was literally a mechanical system — a cam
turning in circles, and at exactly the same time that the piston's at top dead
center and needed to be fired, that little cam pushed the breaker points open.
It's just opening and closing the circuit based on a gear in the distributor.

Tom: In 1975, the federal government
said all vehicles have to be electronic ignition. GM did that. At the time, in
the U.S. market, roughly 65–75% of all sales were GM, Ford, and Chrysler. The
Hondas, Toyotas, Nissans were just really breaking into America in the early
'70s. GM did their own thing — this little hybrid ceramic thick film, which we
do at our plant here in town. Ford and Chrysler did kind of different things —
big circuit boards, big honking things. But at that point, they were electronic
switches. 1975 was really a watershed event.

Tom: The next watershed event was OBD
— On-Board Diagnostics. OBD1, I think, was 1996. You essentially have a
computer, an ECU or ECM, in the car, with sensors hooked up to it. And then
OBD2 — On-Board Diagnostics 2 — came along in the early 2000s. That's the
little dongle underneath your firewall that the technician plugs a cable into
and gets all the information out. That's what turns on the service engine light
that so many people hate because it results in a $500 bill.

Tom: In order to do that, you had to
have more and more sensors. An engine went from essentially no sensors to
pressure sensors, temperature sensors, mass air flow sensors, EGR valve
position sensors, throttle position sensors. OBD2 created all of these sensors.

Government Push vs. Industry Innovation

Adam: How much of it was a push from
the government? Something very similar has happened in the lighting industry
over the last couple decades from the Department of Energy.

Tom: The government was pushing to get
cleaner emissions — the cars were really dirty at that point. When you draw a
line at 1975, some of those automakers might have said, well, we'll get there
on our own speed. So we all need a push sometimes. The government pushing there
was probably a good thing. But then the natural competition kicks in — whose
system is better, produces more horsepower, better fuel mileage, lower
emissions. And then maybe somebody's doing it really well but their costs are
terrible. So it's government part of it and natural competition the other part.

Adam: That's why all the cars from the
early '70s didn't have any horsepower. After the muscle cars died and all the
emissions came in, cars became dogs. Now you have a vehicle with 300 horsepower
and you don't even think anything of it, but the '80s vehicles had 140
horsepower.

Tom: And you can hardly smell the
exhaust out of the exhaust pipe on a 300-horsepower car. They're really clean.
And I don't even think they're close to getting internal combustion engines to
where they could go. There are so many refinements that can be done.

Starting at Wells in 1985

Adam: Now you're at Wells here in Fond
du Lac in an engineering leadership role. From the mid-'80s till now, where
else did you spend your career?

Tom: I started at Wells in early 1985.
I'm from Sheboygan, Wisconsin — local boy. Found a job back home, which I was
looking for. Motorola was a great place, but Phoenix wasn't for me. I have a
big family, so getting back was important. We started at Wells — they were
already starting to make production lines with robots on them. We probably had
a production line to make that first generation GM thick film ignition module.
We had about 15 robots, four of us in that little plant. That big complex right
now was a little 16,000-square-foot building. We had no clue what we were
doing.

Tom: We'd bang those robots into
things. You have to imagine if you have somebody running this all week long,
what types of events might happen? You didn't think of half of them. Somebody
pushed buttons in a different sequence, or a part fell in a weird way when you
tried to pick it up, and the robot arm slams into something. Or the electronics
wasn't that reliable — a thunderstorm would come through and half your robots
would need a new power supply. The early age was kind of the Wild West trying
to figure out robotics.

Tom: My function was — because I did
computers and knew electronics — I was our test engineer. Wells had a handful
of maybe three or four word processors, but most people still did
correspondence with typewriters. I brought in engineering computers and a bunch
of Hewlett-Packard and Tektronix test gear. When a part rolled off the end of
the line, you had to check the six or eight basic functions to make sure it
worked. Programming all those — and there was no YouTube. You just beat your
head against the desk till you figured it out. There wasn't even an industry
colleague you could go to.

Tom: I did test engineering for 23
years. Started out as me, eventually we had a group of five people. Then we
built a reliability lab with thermal shock chambers, vibration tables, and salt
spray. Then for some reason we needed somebody to take over shipping and
receiving, so I put my hand into that too. Wells has always been a
wear-a-lot-of-hats kind of place.

OEM vs. Aftermarket: Two Businesses

Adam: What are some of the challenges
and rewards of being an auto supplier?

Tom: Wells kind of has two businesses
as far as engineering goes. We have what we call an original equipment business
— WEP, Wells Engineered Products. They do OEM jobs for Harley-Davidson, Mercury
Marine, Polaris, John Deere, BorgWarner. That's a collaborative environment
between the Wells design engineers and them to finish up the specification.

Adam: When they come to you, what
percentage of the design is done?

Tom: It's all over the map. Sometimes
you're involved very early, and they want collaboration very early. Sometimes
they know roughly what shape they want and what they want it to do, but they
don't know how to make it do that. For instance, heat dissipation is everything
in a lot of these devices. They don't know if they can dissipate all the heat
in a smaller package. That's a real collaborative effort. We have a lot of
really cool software simulation packages at Wells — FEA, CFD, magnetic
analysis.

Tom: On the aftermarket side, it's
kind of the opposite. You have a part that's already on a vehicle. You're
waiting for the warranty period to end, and then at some point there's going to
be aftermarket sales. For most of its 120 years, Wells has concentrated on
aftermarket replacement parts — maintenance for your vehicle. You're
constrained by the external geometries of that part. It has to fit on the
engine the same way. If it doesn't look almost exactly the same, the mechanic
thinks they have the wrong part and takes it back without even installing it.
So the external constraints are very rigid, but what you do inside to make your
part do the same thing, you have a little more latitude.

Tom: A lot of our aftermarket sales go
to the major retailers. If you drive down the main drag here in Fond du Lac,
you'll see the three or four major auto parts stores. It's a very competitive
market. They all want the best quality parts at the best value, and they want
availability. The “Walmartization” of auto parts — over the years, the major
players that excelled in their industry have taken over a lot of market share.
They needed more parts for more stores, bigger stores with more SKUs. To supply
them, they need diversity of parts, availability, and value.

Offshoring, Onshoring & the Pendulum

Tom: In the 1990s and 2000s, we spent
a lot of time going over to China and the Far East to get lower manufactured
costs. The last five years, the world is a different place. There's some pretty
good onshoring going on, and that's kind of exciting to see the pendulum start
to swing in the other direction.

Adam: We really saw that start when a
lot of the tariffs kicked in back in 2017–2018. That disrupted supply lines
from China. We started to do as much onshoring as possible back then, and then
when COVID hit, it just lit a fire on that trend. Wisconsin has such a great
supply base for mechanical parts. Almost all of our mechanical parts are now
sourced from Wisconsin.

Adam: If you look at the macro
30,000-foot view from the '70s and '80s when there was a lot of domestic
manufacturing — what happened each decade, and what does the pendulum actually
look like now?

Tom: Companies have to adapt to the
changing marketplace, and saying it and actually doing it are two different
things. I've been fortunate — I've spent a lot of time in China. The cool thing
is being in so many different factories. Most of those factories were only
started maybe three years before you get in there. They're not 50-year-old
factories. We look at our own factory and say, if we knew 20 years ago what
we'd be building today, our factory would be arranged completely differently.
But those Chinese factories were arranged to be efficient for their business
right now. They were able to see the wrong way of doing it, start from scratch,
and it's a collaborative process. I came back with ideas.

Low Volume, High Mix: The Core Challenge

Tom: One thing that really struck me —
when we looked at our business from a 1990 standpoint, each vehicle had very
few SKUs. A manufacturer used the same module on five model years in a row,
maybe on three different engine platforms. You'd build relatively few part
numbers but make lots of them. Fast forward 20 years — now you have very many
SKUs, and the number of pieces you have to build for each is very low. We call
that low volume, high mix. But the machinery and equipment we were using to
build them was still made for high volume, low mix.

Tom: What really stood out to me — in
China, they'd have a really small injection mold machine, no bigger than this
table, that might cost $25,000. A tool to make a plastic part on it might cost
$5,000. But we have a $300,000 injection molding machine and a tool that costs
$60,000–70,000. But you're going to make 5,000 pieces a year for 20 years. Do
the math — it's almost a dollar a piece just for the life of the tool. And you
need three injection molding tools and four stamping dies and assembly
equipment and test equipment. The math just isn't there.

Adam: We're a very different industry,
but our value prop is also low-to-medium volume, high mix. As we learned about
tooling, you start to realize how tooling for low volume, high mix also becomes
an innovation constraint. You feel compelled to pay for that tool even if the
original design decision wasn't a good one. What we said is, instead of tooling
up for five components at $60–70,000 in tooling, let's look at 3D printing. For
us it's made economic sense. We can make 60–70,000 parts per year. It's exciting
to think about how, over the next couple decades, as 3D printing and digital
manufacturing happens, how much innovation and creativity that's going to
unlock.

Tom: 3D printing is clearly a game
changer. For us, being under-hood, most of our parts have to survive and
operate to 150°C — 302°F. The 3D-printed materials can withstand that now, but
they also have to have a certain cosmetic appearance and fluid compatibility.
We probably have 8–10 3D printers, maybe even double that, plus a really big
one with about an 18-by-12-inch work envelope that prints down to 0.025
millimeters — about a thousandth of an inch in accuracy. We're knocking out
parts on it all the time. I can't wait for the day when we can print
automotive-grade plastics and not have to buy injection mold tools.

Adam: What we're exploring right now
is metal 3D printers — aluminum. For us in the lighting industry as a smaller
company, the monster companies are taking care of the high-volume product. For
us it's low-to-medium volume, high mix, aesthetics.

Tom: I've seen a couple of shops that
are 3D-printing injection mold tools. The really cool thing — when you're
making an injection mold tool, you have water passages that are bored in with
straight end mills. But they can make the craziest curved, spline, twisted
water passages to control cooling, because when you 3D-print it, it just
happens. They can make a shape that you can't do with a CNC machine. Pretty
exciting.

Wells Engineering Organization

Adam: Just kind of the anatomy of your
engineering organization — what types of engineering positions do you have?

Tom: Besides designing our own
products, we design and build most of our own automation and equipment, and
have all the way back. There's a design engineering group doing new product
development. Those tend to be about 75% mechanical engineers, maybe 25% electrical
engineers. They're working with CAD drafters or CAD designers — all doing
mechanical design, 3D SolidWorks. They work in teams of two to five people on a
project. There's a principal engineer overseeing it and an engineering manager
responsible for multiple pods of those teams.

Tom: Think of two-year mechanical
design degrees from Moraine Park Technical College, Fox Valley Technical
College, Lakeshore Tech in Sheboygan, or the Milwaukee or Green Bay area. The
mechanical and electrical engineers come from MSOE or the UW system or elsewhere.
The competition for good engineering technical people is pretty stiff in Fond
du Lac, Wisconsin.

Tom: For the automation, we have the
same types of people, but they're working on the robotics and automation and
tooling plates. That's a team of about 10. Next to that, there's a test
engineering team — at the end of every production line, there are automated
testers that connect up to the parts and measure everything. They tend to be
electronics engineers who are also computer programmers. That's another group
of six to eight-plus people. And then there's manufacturing engineering — for
machines that are in production and running every day. They're partly
responsible for maintenance, but in a low-volume, high-mix production
environment, they're also assisting with new product launches and getting
whatever new tooling is associated with that line. That's a mix of mostly
mechanical, four-year and two-year degree people, programming PLCs, working
with sensors, pneumatic valves, light hydraulics.

The Production Line: From Circuit Board to Finished Part

Adam: What does the production line
look like?

Tom: In Fond du Lac, most of our parts
are pretty electronics-based. The front end will be an electronic circuit
board. We have surface mount lines — I'd guess five or six big surface mount
lines where you're screen-printing solder paste, high-speed placers for
resistors, capacitors, diodes, transistors, microprocessors, power devices, and
then reflow ovens and in-line testers. So you're making the circuit board
first.

Tom: The other type of circuit board
we do is hybrid thick film. That's a piece of glass ceramic, maybe 25
thousandths thick — about half a millimeter in thickness. Ceramic is much
better with power dissipation, and you get a lot more circuitry in a lot smaller
place. A lot of automotive parts need ceramic thick film substrates. There,
we're printing on resistors and conductors and firing them right in our
facilities.

Tom: Those circuit boards then flow
out to an assembly line. You're gluing or attaching that circuit board onto a
heatsink or a metal back plate, putting usually an injection-molded plastic
housing around it, and then making connections from the outside package —
typically an 8-pin connector and a 2-pin connector for different wire
harnesses. You've got to get from those connector points on your outer package
down to your circuit board. So you're doing interconnects with welding, wire
bonding, and in a few cases hand-soldering of leads that go through a circuit
board, but in most cases automated welds and interconnects.

Adam: What's the largest footprint
part or sub-assembly that you make?

Tom: Think of your iPhone times three
or four. Fairly large. We make fairly large voltage regulators that go on new
Harley-Davidsons. All those things are pretty big clusters with a big aluminum
fin heatsink — they dissipate a lot of power. You think you're just turning AC
voltage into DC voltage, but there are a lot of requirements on one of those
bikes to make sure everything works perfectly — controlling all your lighting,
heaters, everything you need to control, and not pulling too much power off the
engine while keeping the battery charged correctly.

Electric Vehicles: Infrastructure & the Future

Adam: Let's end on electric vehicles.
Where is it going? What are some of the challenges?

Tom: Wells — a little bit. We tend to
be a little more backwards-looking because in replacement parts, we're looking
backwards five to ten model years. But our parent company is very in tune with
when internal combustion engines die, because a big chunk of their business
goes away. There's a lot of thought going into what types of businesses they
want to get into.

Tom: You pull into a gas station with
a Ford F-150 with a 20-some-gallon tank, and in three minutes you can put 500
miles of charge into that, and you drive away. That gas station might have 16
or 24 pumps, servicing 16 to 20 customers with a turnover at each pump of once
every four or five minutes. Even with a quantum leap in charging capabilities —
I don't know how many gas pumps are just in Fond du Lac, Wisconsin, but it's
got to be in the hundreds. How do you get hundreds of charging stations?

Tom: And by the way, if there's an
event that can put 500 miles of electricity into a vehicle in three minutes,
I'm going to be standing a couple hundred feet away from that transaction. The
industry is going to have to figure out how to catch up with charging
infrastructure. But it also has to catch up with how much capacity to build
batteries. There's not enough capacity to build batteries, and then how much
capacity is there to mine lithium? A lot of lithium is in South America,
Australia, some in China. Not a lot in the U.S. If you want energy
independence, you're still going to have to go somewhere else for those
materials.

Tom: Some of the really ambitious
projections are still saying model year 2040 — new vehicle sales will still be
primarily internal combustion engine. It might be a hybrid, but there's still
an internal combustion engine on that vehicle. I just wonder if 20 or 30 years
from now somebody's going to figure out a fuel cell or something different than
lithium-ion battery.

Adam: We do work a little bit in the
charging space. There's definitely a scramble to figure out how to do it. I
don't think there's a unified approach at this point. We're right next to
Sadoff, the recycling company. Just the things that have been built slowly over
decades on how to recycle all the materials inside a vehicle — and all the
dissimilar metals involved with electric vehicles versus classic vehicles —
there's just so many things that have to be worked out.

Tom: Wells is really good at power
electronics. If we need DC-to-DC converters or AC-to-DC converters — every
couple years there's a new gizmo on a vehicle, and we just figure out how to
make them. Reverse engineering something cost-effectively on equipment you
already own is a pretty big challenge, but that's something I think we'd be
really good at. Just like in 1975 and the OBD2 events, Wells just started
making different kinds of sensors. You learn those technologies and leverage
your core competency. You layer new technology into competencies you already
have, and a lot of times it's a really good marriage.

Giving Back & Community

Adam: Anything else? You guys seem
like you're getting more and more involved here in Fond du Lac.

Tom: Our previous ownership groups —
I've worked with them all. But we're much more local-based now. When I say
local, we all live within 30 miles of here, not flying in from a thousand miles
away. Wells has always felt like — people in town would even say, who's Wells?
And we're one of the oldest companies in town — far older than Mercury Marine.
But we've been way too low-profile. So all of us really enjoy getting out,
whether it's at Walleye Weekend or the Envision events, just participating in
the community. Fond du Lac and the Fox Valley in general is just a great place,
and we love being part of it.

Adam: Thanks for all you guys do.
Bringing people through, I've learned a lot. It's nice to turn the various
factories and operations inside out and learn from each other. Thanks for
coming on.

Tom: You bet. Enjoyed it a lot. Let's
talk again soon.

Adam: Sounds great. Thank you very
much.