How Technology Is Impacting The LED Industry (feat. WPG Americas Inc.)

Introduction

Adam: This is the Wisconsin Lighting
Lab podcast. My name is Adam Rupp, and my guest today is Mike with WPGA. Mike,
welcome.

Mike: Absolutely.

Adam: So Mike, let's get started. Tell
me a little bit about your background in the lighting industry and lighting
business, and what WPGA does.

Mike: I started out in the
communications and semiconductor industry. I made a transition into lighting
about seven years ago, and I've been able to utilize some of that background in
an adjacent industry into the lighting space. So I have background in power,
communications, and now a lot of background and knowledge in the LED space as
well. WPG is a components distributor with a focus on lighting applications.

The Rise of LEDs

Adam: Up until maybe ten or twelve
years ago, the lighting business was really separate from the rest of the
electronics business that you had some experience in. With LED, those two
worlds seem to be merging a little bit with solid-state electronics. Is that a
fair statement?

Mike: Yeah, they're merging
incredibly. A lot of the initial development investment by the Department of
Energy was to give some resource money out into the market to move it along.
They needed to hit that power budget — to reduce power consumption in lighting
specifically. And now the lighting space, because of the adoption of LEDs,
there's a lot more technology just around driving those LEDs. So control and
other spaces are all looking at that. It was always the DOE's concept and goal
to achieve power reductions through solid-state lighting efficiencies, but also
through control. So both of them together are going to get them to what they
initially set out to.

The Future of Controls

Adam: As far as the control side of
it, how far along are we in that process as an industry? Different companies
are probably further ahead than others, but in my mind it seems like that has
really just begun. With LEDs, from a lumens-per-watt standpoint or an adoption
standpoint, the technology seems to be leveling off a little bit — you don't
hit those huge gains in efficiency anymore. And now it seems like the next wave
is gains on the control side. Where are we in that controls process — still the
early stages or is it mature?

Mike: Controls were introduced right
along with the LEDs; people just focused on the LED portion. From my
perspective, the industry had a lot to get used to where people didn't really
pay attention to color temperature — everything looked good. Now, fast-forward
over the last 10 years, everyone's very familiar with it and wants to see more
features in their lights. And there have been studies that show how that helps
with productivity in the business place or just at home. The controls were
mostly the focused effort of the bigger lighting tier ones that had the
resources and engineering to work on it. So it's been there for a long time,
but the industry now is at a stage where we have a handle on the solid-state
lighting — the LED, the quality and type of light — and now there's the need
for this control. And the control adds cost savings. DLC is implementing their
service to give you a certification — it allows you to get rebates if you use
controls.

Adam: Interesting. Along those lines,
I've been doing this for about 10 years now. I remember when we first started
selling LED as a distributor — I won't say it was easy to sell on efficiency,
but the ROI was so fast going from a thousand-watt down to 300, or 400-watt
down to 120, where the customer would get such massive gains from LED only, it
was sometimes challenging to sell them on a complicated control system to get
that extra 5 or 10 percent. And on the solar side of things — solar has been
around for a long time — we'd have customers come to us and say, hey, I want a
solar solution, and we didn't have any solar options. Part of what we'd walk
them through was: with this one product, this one fixture, you can eliminate 70
or 80 percent of your energy costs. To get that extra 20 percent in a solar
solution, you incur a lot of maintenance, the system gets more complicated. I
see controls now where that's the next wave — the controls cost has come down,
and from a contractor expertise standpoint, there's people that have more of an
understanding of how to manage the control side — I call it the contractor Geek
Squad. So the market is now primed for that.

Mike: Yeah, and that's what I'm
looking at — not only what's happening in the lighting space, but things in the
adjacent industries that people in the lighting area will take and make use of.

Adam: Are you seeing telecom or what
other types of industrial or commercial products?

Mike: The one that's having probably
the most impact is IoT. That's been adopted by a lot of industries outside of
the lighting space, and what they're trying to do is make controls
interoperable. Any time a company brings a product to market and is dependent on
another company for it to operate, that's a point of concern. The idea of
Bluetooth Mesh SIG is that there are a number of companies making those chips;
if they're all SIG-approved, I don't need to worry about some semiconductor
company disappearing on me.

Adam: So more or less in the lighting
space, for whatever reason, you have people running in a lot of different
directions, whereas in consumer products and retail products there's more
standardization around those IoT products.

Mike: Yeah, IoT is going to bring — or
is attempting to bring — that standardization that's developed in other
industries and marketplaces to lighting. The technology out there today was
pretty proprietary. You've seen it out there, but you pretty much have to
depend on that company and hope they continue to support you. There are always
issues with legacy products not interoperating with the new products they bring
out. It's a struggle.

Adam: Right. We had Synapse, a smart
lighting controls company, on the podcast a few weeks ago, and their system
relies heavily on ZigBee. They more or less have a ZigBee application that
they've built on top of the ZigBee protocol, and they seem to do a really good
job with non-video, non-data. Is it true that the bandwidth in ZigBee is much
lower than Wi-Fi or Bluetooth? What do you see as the competing technologies
and functions?

Mike: What's really nice out there in
the marketplace is the amount of technologies that have been developed, and I
don't think any of us think it'll be one or the other. The transport mechanism
could be wireless or wired. There are a number of companies out there that have
power line communications solutions over AC.

Adam: Really?

Mike: Absolutely. They were early
adopters targeted for meter reading, and now there are a number of lighting
OEMs and manufacturers that have introduced products where you communicate over
the AC. So the AC has to both power the fixture and also communicate with the
network. That's basically what you're looking at with power over ethernet now —
DC-based distribution. So there's communication for DC and also for AC, and it
can go up to 10 megabits or higher. In communications there's always a
trade-off: the higher your bandwidth, the more susceptible you are to
interferers or other noise. So the slower speeds are a bit more robust. If you
look at the wireless LoRa system, it's a very low data rate but it can
broadcast easily 20 to 25 miles line-of-sight.

Adam: Wow, that's amazing.

LED Chip Technology

Adam: I want to get more into the
controls stuff and IoT, but I want to talk a little bit about LED chips
themselves. LED technology has been in the US market for a while now, and when
we first got into it, there were major gains in efficiency and lumens per watt
being picked up. I remember when a 400-watt metal halide replacement in LED was
this big — now we have a seventy-five thousand lumen fixture we're coming out
with that's way smaller than the original 400-watt replacements. Do you see the
efficiency gains leveling out on the chip side? Have a lot of the resources at
the different chip companies been directed elsewhere, or are they still pumping
investments and resources into making LED technology more efficient?

Mike: I can draw a parallel to the
efforts I see in this space and other semiconductor industries: as long as
there's a market and people can realize enough margin in their product, they'll
continue to invest in the fabs and process improvements. All of them right now
still have a lot of process improvements, and there's also equally large
investment in packaging technology. So you're developing your fabs and your
wafers, improving that, but you're also having to adopt new packages — things
like flip-chip or other technologies that help make it easier to integrate into
the fixture architecture. You look at some of the people out there with 220
lumens per watt — that's going to continue to creep up as people still want
more photons for less money. Basically, I look at it as we're delivering
photons.

Adam: Yeah, it's hard to argue that.

LED Technology Beyond Illumination

Adam: As far as LED technology being
deployed to end markets other than general illumination — I know there are
companies coming out with LED products that disinfect operating rooms, products
with RGBW, sending data over LiFi. What are some of the cool applications you
guys are seeing in the market that people are using beyond just making a
fixture light up?

Mike: For LEDs as the light source, I
think everyone right now is able to adopt some of this new improvement because
it's evolutionary, not revolutionary. That's a good way to put it, and it keeps
their products up with the rate that the LEDs are improving. In some spaces, it
doesn't necessarily cause a huge dramatic shift in the product they release
this quarter to the next — there's going to be some improvements and cost
improvements. They may just add a new product that the customer can take
advantage of. So I'm looking at the revolutionary things that are happening.
Right now on the LED landscape, there's a new packaging technology people are
very interested in and exploring, and that's probably the most revolutionary —
flip-chip and surface-mount LEDs.

Adam: What about healthcare and other
markets? I see LEDs being used for red light therapy, for skin problems, for
disinfecting things or treating other ailments. Have you guys been involved
with any customers doing some of those more cutting-edge use cases?

Mike: Yes, we have been looking at the
UV space, and I'll say that UV LEDs have some tremendous hurdles to overcome.
With UVA — you get into deep ultraviolet, in the 400-nanometer range — you can
build a product that can do disinfection. That's really at the edge of the UVA
space; that energy, that spectrum, isn't so energetic. But if you get out into
the 250-nanometer UVC or UVB, the LEDs are really struggling. They remind me of
what LEDs in 2008 looked like — about 10 lumens per watt. They don't last long.
A lot of material science needs to go into that to come up with packaging that
won't self-destruct.

Adam: So more or less, the packaging
we're coating them with — is it additional phosphorous? Are they using other
material sciences to dial in that radiation wavelength? Like we have on the
back wall here, we have different LED chips from turtle-friendly up to 7,000K —
is it the same technology to tune in those different colors, or is it a
different material science altogether?

Mike: For us, we're in the visible
spectrum, and that's important to us. So we use phosphor to create a broader
spectral distribution. UV and IR — in horticulture and similar applications —
is monochromatic. You're just at one specific wavelength, so there's no
phosphor. In fact, that's the best way to deliver the most photons. Typically
it's just a clear coat over the die for protection — you'll just have a window.

Adam: So when you have these LEDs that
aren't very efficient, like the old visible ones, you have a lot of heating?

Mike: Yes, many photons, and as you
said, they self-destruct. The energy, the frequency, is destructive. That's why
you can't leave things out in the sun — it breaks down a lot of materials. A
lot of material science will have to go into the packaging. You can't use the
same silicone you package with the visible LEDs.

Adam: Cool, that's some interesting
stuff.

Drivers & Power Supplies

Adam: Switching gears now onto drivers
and power supplies — you come from the power business. One thing we've been
doing a deep dive on the last 12 months or so is trying to understand exactly
why in some cases drivers might not last. We work on a lot of applications that
aren't friendly to power supplies — could be a 120-foot-high mass light pole
that turns into a lightning rod, retrofitting LED technology onto grid
technology that's several decades old, in addition to surge protection. We try
and select the highest quality power supply we can. What do you guys see power
supply vendors doing to create more robust packages?

Mike: It's always a trade-off — cost
and performance. There's still the ongoing decrease in the cost of LED drivers,
and most of them are AC to DC. What we are seeing are more companies looking at
adjacent markets. An influencer would be that our infrastructure in this
country is antiquated and overloaded — so everyone's discussing smart grids. If
you look at renewable energies, windmill farms are distributing power via DC,
and in South America, hydro dams have long DC runs. So a lot of lighting
manufacturers are doing remote-mount drivers. If you can reduce what you need
to do in terms of putting protection on every fixture — if it has an AC-to-DC
converter — and instead move that back and put better protection on the one big
power input upstream, isolated, then you're not putting redundancy and cost
into every fixture. We see some companies remote-mounting just because of form
factors they have to meet; it's not always possible to get drivers into the
fixture. We're working with a supplier on a DC-to-DC converter that's six
millimeters high, 50 watts — a simple buck converter. So there are ways to
solve a problem without looking at the way you've been doing it.

Adam: Yeah, to get there you really
have to understand the whole architecture — what the fixture is being plugged
into. Not just focusing on the output, not just the light, but going upstream.
Even a checklist for the contractor: is the utility transformer outputting the
actual voltage that it says it should? Is the distribution panel up to NEC
standards with proper surge protection? As a factory, we don't always have that
engagement with the local people installing or interacting with the products.
But for some of these mission-critical applications, we do have that engagement
with engineering teams — trying to understand their entire system versus just
the hot, neutral, and ground the fixture is being plugged into.

Mike: It's interesting you bring that
up, because that's one of the features that a couple of suppliers are putting
into their products — the ability to monitor the incoming line and the output,
and having a mechanism so you can —

Adam: I think that's huge when we talk
about controls and power monitoring. Part of the problem is, if there's an
issue in the field, it's getting everybody on the same page. A lot of factories
will say, ship the product back, we'll send it out to our labs, we'll do an
analysis, figure out what component failed, identify whether it was a lightning
strike, some other upstream problem that caused it to fail, or a part defect.
More often than not, it's some type of over-voltage problem. But if there's a
monitoring device on board — or even at the bottom of the pole — that can say,
hey, this thing saw 20,000 volts, or it sees 20,000 volts every time the
laundromat three blocks away starts up, that at least gets the messaging and
gets people moving in the same direction. That's an important problem to solve.

Mike: Yeah. All of us having been in
this business — a lot of the installs are where you have most of your issues.
Trying to understand it from a return, when returning one component — very
difficult.

Adam: It is, it is.

Smart Lighting — The Rise of “Computers”

Adam: Back to smart lighting and
controls. One of the things we talk about a lot is, two or three years from
now, we'll be way more of a computer manufacturer than we are a light fixture
manufacturer. What do you see as the tipping point in the market where light
fixtures go from their primary function being producing photons to a secondary
— or even third — thing people think about when they buy a product? And what
has to happen before that happens in the space?

Mike: A good example that exists out
there today is building control, building automation, which is the farthest
along. There are building automation companies adding a lighting component to
their product offering. For lighting, it has a different cost point in the
industry just because of the nature of what it is, so there has to be something
available for lighting that doesn't add too much cost — where the OEMs and the
installers can afford to put it in. In that case, the computer, the processor,
is at the building management level, and your sensor suite is at the fixture
level. Because if you start to have redundant computers in a warehouse, it's
going to get pretty expensive.

Adam: Exactly.

Mike: And you don't necessarily need
some of the roles that you may need in another industry. Some of that does need
to be looked at — HVAC needs to be monitored a lot. If you're not going to do
data auditing reports for your fixtures, then you don't want to buy a lot of
overhead. And if you want to connect that into the cloud — the IoT space is
addressing both of those. The nice thing is that they're entering with more
cost-effective solutions and trying to reduce the complexity. I think that's
the kind of learning curve that's going on — how do you commission these
products and set them up?

Adam: Absolutely. Who does that?
Everyone's software is different; there's not really one expert. And we see
that a lot on the sports side of things. We had some contractors and
distribution and agency partners up here last week, and we were showing them
the next rollout of our product design. In the contractor's world, they want as
much done at the factory as possible. They want to install the sports light on
the existing cross arm on the existing wood pole, hit the Go button, and have
the thing work. If there are zones that are needed, or packages for light shows
or other things, they don't want to be programming in the field. So that's
where we're really looking — if we can solve problems upstream at the factory
and make the installer's life easy, that's good for everybody.

Mike: Yeah, I can see why you think
you're going to be in the computer space, because if you're offering those
products to the market, you have to be the one that provides that service.

Adam: It's like the enterprise server
space — you know, Amazon has a 22-megawatt server farm down in Tennessee. The
companies that sell those servers — rows and rows of servers — they have the
sale of the product, but then there's a post-sale conversation about
maintenance. In those cases, is it still the electrical contractor doing the
install? I know there's a voltage range where it's more of an IT professional
versus a contractor, but I also think that sometimes muddies the waters a
little from a legal standpoint. A licensed contractor has to install the AC
products, but with a server farm, in theory, it's all low voltage. So it's a
different market, a different skill set. Is my understanding correct on that?
Who does those installs?

Mike: It is low voltage, but in the
industrial low-voltage space, low voltage is up to a thousand volts.

Adam: Oh, wow.

Mike: So data server farms have
actually been using three- and six-hundred-volt distribution for quite some
time. That's a pretty efficient way to transport a lot of power. You do need
electricians there, and they had to overcome things like DC arcing, which starts
to happen around three hundred to three hundred and fifty volts. So there are
insulated wires and connectors — that's mostly what'll burn out from the arcing
that occurs at those voltage levels — and they've overcome all that.

Adam: Mike, is your time
lighting-specific? It sounds like you have really good insight into other
spaces. I know your company works on pretty much anything electronics, but is
your role focused on lighting?

Mike: It is. Like I mentioned, we're
vertically focused on a couple of markets. I'm involved in the lighting
segment, so I take all that previous knowledge and see where we can apply it
into the lighting solution space.

Adam: That's awesome.

Driverless LEDs & Remote Power

Adam: I know you've got some notes. Is
there anything you want to go through, any topics we should touch on?

Mike: Just back to remote power. One
of the things people are doing more and more — if you do a remote mount, you do
a driverless fixture. So the power at the fixture is changing: no driver, no
power supply on board, where you just run 48 volts or something less to it. Or
they're putting maybe just the DC-to-DC in there, or they have AC running to a
circuit on board, driver on board. I see people experimenting a lot with
various technologies that are available, and maybe it suits their applications
for what they need.

Adam: If there's a DC-to-DC power
supply on board, isn't that still a failure point at the fixture that would
have to be serviced?

Mike: Right. If you look at AC to DC,
that's a much more complex device. What makes an LED driver different from a
power supply is that it does need to meet some lighting specifics — power
factor correction, and you typically have a dimming circuit. All that adds
money. If we just put a simple buck converter on there, that's really a small
footprint, bill of materials is much lower cost, better efficiencies, and it's
not having to deal with AC anything. It becomes a nice, convenient little
driver. It gives you basically what you're looking for, which is tight
tolerance control — you need constant current output.

Adam: Is that more or less stepping up
the voltage at the fixture, or bucking down?

Mike: Bucking down. Just buck down and
keep it a very simple circuit. So I was looking a lot at the DC space — and in
automobiles, the issue they're struggling with is they have 350 pounds of
copper wire on a car today, almost five miles of wire.

Adam: Sounds expensive.

Mike: Yeah. There are companies that
are very big in that space making these wire harnesses, and it's getting so
complex out there with all the automotive manufacturers — you don't really have
one harness for everybody, you have thousands of harnesses. So they want to
reduce the wires, just from a manufacturing standpoint, and installation needs
to be simpler — hooking these harnesses into the vehicles is also a challenge,
especially when you have to switch over to another model. So they're looking at
power line communications and DC. They've developed some very robust power line
communications for DC to reduce that wiring and put the communications and
control into the power grid of the vehicle. They just need to make a big giant
integrated circuit.

Adam: Yeah, they are. In fact, there
are institutes in Germany — material-based and research centers — doing
interesting things with the automotive industry to actually inlay conductive
traces into the body of the vehicle. So on the door panels and the fender panels,
they can put down a groove with a plasma cutter, lay the material — whatever it
is, hardened by UV — and then it becomes conductive, and there's no wire.

Mike: So what happens if there's an
accident and it breaks the circuit?

Adam: Well, yeah — maybe they should
put it in the carpet on the floorboard or whatever. That's interesting stuff.
Well, cool. What else? Anything else we want to run through?

Mike: No, I would just leave it that
it is very much changing out there with controls. I think the IoT space has
really helped move everyone that's been doing controls one way to start
considering more industry-standard ways of doing it.

Adam: Awesome. Exciting times.

Mike: It is.

Adam: Alright, well, thanks, man.
Appreciate it.

Mike: Yep, thank you.