June 19, 2025 • 32 mins
Recycling plants take in a huge amount of random (and occasionally hazardous) stuff, which they then have to turn into reliable outputs that their customers will buy. That’s why Rebecca Hu Thrams calls recycling “the most demented form of manufacturing on the planet.” Rebecca is the co-founder of Glacier, and her problem is this: Can you use AI and robots to make recycling a somewhat less demented business?
Get early, ad-free access to episodes of What's Your Problem? by subscribing to Pushkin+ on Apple Podcasts or Pushkin.fm. Pushkin+ subscribers can access ad-free episodes, full audiobooks, exclusive binges, and bonus content for all Pushkin shows.
Subscribe on Apple: apple.co/pushkin
Subscribe on Pushkin: pushkin.com/plus
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:15):
Pushkin. Imagine you're running a factory. You've got to put
out a consistent, high quality product that your customers will buy,
but you have no control over the raw materials that
come into your factory every day. One day, half of
what comes in is literal garbage. The next mixed in
(00:38):
with your usual inputs is some random lithium ion battery.
It's a fire hazard and also a surfboard weird. And
you have to deal with all this stuff and still
keep getting your product out the door. This is literally
the way the recycling business works. Recycling plants take in
(00:59):
a largely random, occasionally hazardous stream of stuff, a stream
of stuff that changes in a pretty unpredictable way from
day to day, from hour to hour, and then recycling
plants have to turn that random stream of inputs into
aluminum and plastic and cardboard that other companies will buy
and use to make new stuff. This is why my
(01:22):
guest today calls recycling the most demented form of manufacturing
on the planet, and it's why she and her colleagues
are trying to use technology to bring some order to
the recycling chaos. I'm Jacob Goldstein, and this is what's
your problem the show where I talk to people who
(01:44):
are trying to make technological progress. My guest today is
Rebecca who Troms. She's the co founder and CEO of
a company called Glacier. Rebecca's problem is this, how do
you use AI and robotics to make recycling a somewhat
less demented business. If Rebecca and her colleagues are successful,
they'll not only help recycling plants work better, they'll help
(02:06):
companies figure out how to recycle more of this that
they're sending out into the world in the first place.
Our conversation started with Rebecca talking about the moment around
five years ago when she and her co founder decided
to start the company.
Speaker 2 (02:21):
I was, you know, even at the time, obsessed with trash,
just obsessed with where does all of our stuff go?
And it's one of those I call it like a
Matrix moment or a red pill moment where once you
realize that you've never thought about where all of your
garbage goes after you put your bins on the curb,
you can't unsee that, right. And so this was also
(02:42):
around the time where there was a lot of change
happening in the recycling industry. So we're rewinding to roughly
twenty eighteen twenty nineteen. One cataclysmic shift for the industry
is that China, who had previously been the world's largest
buyer of recycled feedstock to make into new things. They
basically said very rapidly, you know what we're taking in
(03:03):
the world's recycling, but most of this is trash. People
are not sorting it well enough. We're getting a ton
of contamination, and we don't want to end up as
the planet's landfill or incinerator. So we're going to drastically
increase the bar on quality of what we're accepting. And
then that caused shockwaves throughout the globe and certainly for
the US, where suddenly recyclers for the first time in
(03:23):
a long time, were like, the game is not to
just crank through all this recycled material, bail it and
ship it overseas. We actually need to invest a lot
more in increasing the bar on quality and on purity rate.
And what's even more challenging is that a lot of
the backbone, historically and even to this day, is still
just people standing next to conveyor belts sifting through our
(03:46):
recycling in our trash and of course that's not only
very dangerous, it's not a very well compensated job. There's
a lot of hazards there. But also there's this massive
sort of labor shortage.
Speaker 1 (03:57):
Yes, it seems like a robot friendly moment, a robot
friendly environment. So like, what's your move, what's your first move?
Speaker 2 (04:03):
So my co founder had already at the time, he
wasn't my co founder, he was just a friend of
a friend. He was already pretty intent on this idea
that hey, you know, automation AI, all of these technologies
are so good now that for the first time, we
could feasibly rapidly commercialize a purpose built industrial robot, specifically
(04:24):
for recycling sortation.
Speaker 1 (04:26):
Huh.
Speaker 2 (04:26):
And it's not going to cost us massive amounts of capital,
and we can actually do it in a matter of
like a couple of years.
Speaker 1 (04:33):
So this idea that oh, now is the moment? Is
it computer vision? Like, what is the underlying technology that
made five years ago or whatever? The moment when oh,
we can do this in a way that hasn't been
done before.
Speaker 2 (04:45):
Yeah, Honestly, it's the confluence of a lot of things.
So I'll break it down into sort of the computer
vision piece and then also the hardware or the robotics piece.
When you think about where industrial automation has come from
even to this day, a lot of those technologies are
operating in really well defined, truly repetitive roat environments. So
(05:06):
think about a robot at a warehouse, and it's literally
just palatizing identical boxes over and over again. And so
to harken back to this idea of recycling plants as
being this extremely volatile manufacturing environment, even if you have
automation that's just sorting let's say aluminum cans, right, You're
talking about aluminum cans that your computer vision needs to
(05:28):
detect in infinite varieties, not just thinking about the wide
variety of cans that are on the market in all
of their colors and designs, but also the fact that
they show up not as pristine cans but crinkled in
various ways, stuffed into bags. Like, there's so much heterogeneity
that even just identifying that item on conveyor belt that,
(05:50):
by the way, has dozens of other types of items
on it is already a massive challenge that only recently
has been something that we can adapt to in a
cost effective way. And then you can layer onto that
the fact that now you're not only seeing those items
with this computer vision system, but you also need to
find a way to actually go and grab that material
(06:10):
and sort it into the right location. So when we
talk about advances in sort of off the shelf parts
that you used to design and make your robot, or
even like the gripping technology that's available, a lot of
that even a decade ago, would have required an immense
amount of r and D, with a much bigger team
and a much higher price tag to get to the
same point that we've gotten to after just a couple
(06:32):
of years.
Speaker 1 (06:33):
Tell me about the first one you built. Tell me
about building a prototype and putting it in the world.
Speaker 2 (06:37):
Man, So there are a lot of different stages to
our prototyping. The first prototype, if you want to go
way back, was when I hadn't even decided to start
this company with my co founder yet, but I did
tell him I would help him learn about the industry
see if there was some sort of a business to
be had, and we met in his kitchen in San
Francisco where he had a little corner set up. So
(06:59):
like very very simple. I think that it actually involved
a used yogurt tub as like this rotating wheel with
a piece of string tied around it, Like that's how
janky we were talking, but it worked and we were like, Okay,
there's something here. The first piece of equipment we actually
installed into a recycling facility also tells you a lot
about the constraints that these facilities are under. I was
(07:21):
kind of thinking we had to have this super built out, sophisticated,
polished thing that we've proven out to the nines in
the lab, and we actually called up a number of
recycling facility operators nearby and one of them was like,
you know what, when you got something, just like bring
it in here and try it out, because literally I
remember him saying, if your robot can pick one more
(07:43):
can than I would have gotten otherwise, Like it's already
worth it to me.
Speaker 1 (07:47):
Just try it. Was there a rat moment, tell me more,
did you see a rat?
Speaker 2 (07:54):
Literally?
Speaker 1 (07:54):
I was like, what is that not a metaphor not
a metaphor a rodent?
Speaker 2 (07:59):
I have seen many rats. Literally, Even on that first
install there was a moment where I was like literally
army crawling under a conveyor belt to fasten one of
the legs of the robot, and I came eye to
eye with a rat, who then, of course ground a
piece of food that was on the floor and then
scurried away. Right, there are many friendly critters running around
some of these facilities.
Speaker 1 (08:19):
Was there any part of you that kind of loved it?
Speaker 2 (08:22):
Oh, one hundred percent.
Speaker 1 (08:24):
So let's talk about where you are today, both on
a kind of micro level, like what your robot or
robots look like, and then also a little more macro
of like the scope of the business. Do you have
like one basic robot what to look like or you
got a bunch of them?
Speaker 2 (08:38):
Yeah, so we have one base model of robot. We
actually are already working with several dozen customers across the country.
But if you imagine any sort of conveyor belt in
an industrial facility, our robot think of it almost like
a table, right, So it's got four legs and it
kind of sits over that belt, and then the guts
of the robot or the mechanisms doing the picking are
(09:00):
kind of over the top of that conveyor system. So
you've got these arms that are going back and forth.
They can pick up something from the belt and then
they carry it off to one of those sides of
the belt. Where they actually drop it into the right location. Okay,
so that's the robot. And now one other thing we
haven't really talked about is this computer vision system.
Speaker 1 (09:19):
So that's yeah.
Speaker 2 (09:20):
Imagine basically like a camera with some lights to illuminate
the belt sitting on this little rig that's a little
bit upstream of the robot. So the material passes under
the camera, the camera has a second to process or
a should I say, a couple milliseconds to process what
it's seeing, and then not only does it tell the
robot you know, hey, there's a can coming in this
(09:40):
location picket and then put it into this other spot,
but what we're also finding is that that data as
a standalone is also able to massively advance these operator's
abilities to understand and optimize their facilities. So that's opened
up a whole new world of use cases.
Speaker 1 (09:58):
Because they weren't gathering data in that kind of way before.
They only sort of knew what was coming through in
a very gross macro way.
Speaker 2 (10:07):
Right, And this gets back to you know, recycling facility operators.
I have so much admiration for how they have gotten
really resourceful with trying to understand their operations. But you know,
to give you a sense of things. The state of
the art in the industry to this day is still
mostly manual audits. And when I say that, I mean
imagine taking a half ton of material off your line
and then literally having two to four people hands sort
(10:30):
and categorize and wigh each item to understand what's coming through,
and then assuming that that half ton is representative of
like the thousands of tons coming through your facility on
a yearly basis. I often tell the story of one
of our early data customers, this gentleman who runs a
recycling facility in California. When he met me, he was
telling me that he had mounted a goat pro camera
(10:52):
above his conveyor belt, the one that was basically supposed
to be all trash leaving the facility, but he knew
he was missing some good stuff, and he would spend
an hour a day after work going frame by frame
through some random snippet and manually tallying how many cans
and bottles were on that line, and then using that
to back into what he would try and change in
(11:12):
his operation the next day, and then he would check
that day to see if it changed anything. And so
imagine his delight when I told him, hey, actually we
can mount our own camera on there, and suddenly we'll
just give you access to a dashboard.
Speaker 1 (11:24):
A machine will literally count everything that goes.
Speaker 2 (11:27):
Literally in real time. So we're seeing that, you know,
the robot is this incredible foot in the door with
a lot of our facility partners, but that everyone's starting
to realize that, hey, actually this data can also help
us understand the entire world, not just the location where
the robot is sitting either.
Speaker 1 (11:43):
So at this point is it kind of a robot
business on the front, but really you're like a computer
vision data AI business.
Speaker 2 (11:52):
So a lot of customers come to us saying, you know,
I literally had a gentleman tell me a couple months
ago like I would never pay you to tell me
what I already know about my trash. And I'm like,
I'm not going to convince you that you don't already
know everything about your trash. But you know you want
a robot, Let's get you a robot. And that has
sense of all the conversation where we're just sort of
starting to show him this data and he's like, oh,
(12:12):
actually I didn't realize that this was the case. The
flip side is also true where someone's like, I don't
know if I need a robot yet, but I'm really interested
to see what I'm losing on the back end. We
install that camera and then suddenly, lo and behold that
data makes the case that, holy cow, I'm losing so
much stuff. I don't just need one robot. I maybe
need two or even three robots. Right, So it's kind
of this mutually reinforcing flywheel that's been really integral to
(12:33):
the success of our business so far.
Speaker 1 (12:36):
Tell me about your work with Amazon and with Colgate Palmolith.
What are you doing for them with them?
Speaker 2 (12:41):
Yeah, So to start, maybe just I'd love to explore
this idea of what is a circular economy because it's
a buzzword that gets thrown out a lot, but it's
really important to understand why Amazon and Coliate and their
peers matter here. Right now, we are living in mostly
a linear economy. In other words, someone makes a thing,
we consume a thing, and then we dispose of it. Right.
(13:02):
A circular economy tries to turn that process into a circle.
So instead of throwing it out in a landfill forever
or incinerating it, that material gets brought back to the
front end and reused somehow to make new stuff that
we can then consume. And the ideal is to make
this go on forever so that we limit our resource consumption.
So now that we have this growing base of recycling
facilities that are you know, gathering data, that are getting
(13:25):
a better understanding of what's coming into their facilities, what's
actually being bailed and sent out to the end markets,
We're working with companies like Amazon and Colgate on a
number of fronts. You know, the first is even just
to understand where is all of that packaging going to
their credit They and several of their peers have realized
(13:45):
that there's a paradigm shift possible now from we have
designed a thing that's technically recyclable, you know, our packaging
R and D engineers have made this awesome mono material
HDPE toothpaste tube in Colgate's example, to now trying to understand, okay,
well we made this thing that is recyclable, is it
actually getting recycled? And that was a lens that we
(14:08):
couldn't really get at scale before and so now with
Glacier's technology, we're able to monitor in real time, you know,
how much of these tubes are actually ending up at
the recycling facility, and once they're in the facility, are
they ending up in the right place, are they being
sorted correctly such that they can actually be turned into
new stuff, or are they ending up in the landfill,
(14:29):
in which case you know, suddenly this recyclable tube isn't
very recyclable at all. So we're starting to answer these
really really critical questions.
Speaker 1 (14:37):
So Colgate knows whatever how many tubes of toothpaste they
sold in a city, and if you are working in
the recycling facility for that city, you can actually count
how many tubes of that toothpaste came down the recycling
line and how many tubes of that toothpaste wound up
in the bin. I mean, is that the reductive version
of what you're.
Speaker 2 (14:54):
Saying, essentially? Yeah, And where we're even seeing now, like
with these rapid advances in AI and in detection, that
first of all, it's no small feet to even define
what is a tube and how do you tell the
difference between a toothpaste tube versus a sunscreen versus a
lotion tube. But now we're getting to the point where
we can actually say the brand of toothpaste, it is
(15:16):
like from all of those visual markings, and so we're
just seeing this sort of Cambrian explosion of interest from
a wide variety of different you know, packaging producers and
brands to really start understanding this previous black box on
what happens once they release this packaging into the wild
for consumers to buy.
Speaker 1 (15:34):
So, I understand that California has a law that is
in some fashion supposed to put companies like on the
hook for their for their products right after they're used,
to incentivize companies to have their products be recycled. Right.
Is that am I characterizing that law right? And is
it relevant to your business?
Speaker 2 (15:53):
Yes? So I believe you're referring to EPR or extended
producer Responsibility laws. For those who may not have heard
of EPR, it's essentially this premise that, you know, if
our recycling and waste system is supposed to find a
way to do something good with all the stuff throwing out,
the people making all the stuff that we're throwing out
should probably have some skin in the game to make
(16:14):
sure that that stuff gets either disposed of or reused properly. Right,
And so EPR laws are already in effect throughout Europe,
throughout Canada, some other regions, and then they've been passed
in a number of states in the US, including California.
Now while EPR in the US is still in its infancy.
In other words, it's been passed in a number of states,
(16:36):
but there's a lot of hairiness to figuring out how
to actually implement the system across all the producers selling
into a state and all of the recyclers operating in
that state. It is, I think a step in the
right direction because in a lot of ways it helps
to create that circle we were talking about earlier. You know,
you're seeing that a lot of brands and producers are
(16:58):
starting to take even more of a vested interest in
understanding what is happening to all of their packaging, because
they know that imminently they're going to need to start
proving the sort of end of life outcomes for that
packaging in order to you know, one, not be heavily
fined and then two maybe even have a right to
continue selling into that state.
Speaker 1 (17:17):
It seems good that Amazon and Kolgay Palmolive are trying
to figure out if the things they make that are
recyclable are actually being recycled, but it seems like for
that sort of thing to happen at a meaningful scale,
you would need laws basically, right. I mean, if the
companies are just incurring the cost either out of the
goodness of their heart or in the hopes of you know,
generating goodwill that will lead to higher revenues, those seem
(17:41):
like marginal cases. Are the EPR laws such that you
think it will become a meaningful part of your business,
a meaningful part of the world, that companies will in
fact be on the hook to figure it out, or like,
what do you think is going to happen?
Speaker 2 (17:53):
You know, I will say that early indicators are that
all of these states are taking it quite seriously. So
in addition to requiring a lot of these brands and
producers to pay into a massive fund upfront to even
just start implementing some of this movement, a lot of
these states are also you know, we're seeing that some
of the kind of like early deadlines and fines for
(18:15):
non compliance are actually being upheld, which I think is
a really strong signal to the market. Hey, this is
something that needs to get taken seriously. Now to your point,
I do think that at the end of the day,
whatever flavor this legislation takes. The key to make sure
that recycling is still a viable and sustainable value proposition
(18:37):
is that there needs to be some sort of an
end market for that material, right because let's say these brands,
even if they're required to pay billions of dollars into
this EPR system, if there's no one on the back
end to receive that material that these recycling facilities are sorting,
then recycling can't really happen. At the end of the day.
Speaker 1 (18:57):
Someone needs to buy the bail of.
Speaker 2 (19:00):
Plastic exactly exactly. But if they can have guarantee that
there is a buyer on the other side, right that
that person or that company will buy at a certain price,
then suddenly they can sustain that business quite well for
the long run. And so to that point, you know,
one other model that's often brought up in the realm
(19:20):
of legislation is actually minimum recycled content laws, because it
kind of gets at the same issue from the other side,
where you say.
Speaker 1 (19:28):
Basically creating demand, creating demand for a bale of recycled
plastic coming out of the recycling facility.
Speaker 2 (19:35):
Exactly, and it kind of disentangles the market for recycled
feedstock from the market for virgin feedstock, which is another
great way to kind of catalyze the movement of that
material throughout that recycling ecosystem.
Speaker 1 (19:52):
We'll be back in just a minute. What are you
trying to figure out right now? What's a big thing
you're trying to figure out?
Speaker 2 (20:07):
We are at a really exciting inflection point, a glacier
because of I think two big things here. The first
is just how do we scale smoothly and rapidly. You know,
we've gone from a year ago we were making maybe
one robot every three months, and now we have the
capacity to make three to four robots per month, and
(20:28):
we're expecting to go even faster by the end of
this year in the next six months. And then the
other big frontier for us, in addition to just you know,
how do we scale and get more of our stuff
out there, is what the heck do we do with
all of this data? Right? We've already seen that the
early use cases for this information people are taking to
really in droves. But there's a much more built out
(20:51):
version of the data platform where we say, you know,
we don't just have a camera in one or two
points throughout your facility. We can actually get sensors to
sort of blanket the facility, and in that regard, we
can take a huge step towards becoming more like that
manufacturer who has information on every single step of their
process and respond in real time and know exactly what's
(21:11):
going on at each stage.
Speaker 1 (21:13):
So, I mean, let's just talk about the recycling business
for a minute. The facilities you're talking about, they're just
private companies.
Speaker 2 (21:18):
The vast majority of them are. So this is a
common misconception about recycling, is that you know, these are
all you know somehow like public entities buy our estimation
about eighty or eighty five percent of these recycling facilities
are privately owned, and they can be anything from a
family owned business all the way up to the massive
waste companies like Waste Management, Republic Services, Waste Connections. These
(21:41):
are publicly traded companies that also own many of these
recycling plants as well.
Speaker 1 (21:46):
And the recycling plants are buying recycling from municipalities like
they do they pay for whatever tans and plastic jugs.
Speaker 2 (21:57):
It's actually a very interesting question. It depends a lot
on the condition of those end markets. We talked about
so in today's climate where those markets are really volatile
and a little bit uncertain, oftentimes, you know, these recycling
facilities will get paid by municipalities in order to take
any process that material. But what's interesting is, you know,
(22:17):
back during the heyday of recycling, when you know, China
was buying everything, there was no shortage of you know,
appetite for that material. The equation flipped.
Speaker 1 (22:27):
My sense is some recycling is quite efficient and a
good business, and some is not very efficient in a
bad business. Right, give me the like stack ranking for
recycling best thing to recycle to worst.
Speaker 2 (22:38):
Yeah, the I mean, at the end of the day,
the best things to recycle, according to a recycling facility operator,
would be the things that most reliably will make you
the most money. So top of the stack would be
aluminum cans, because there's always a market for those. They're
super easily recyclable, and to recycle an aluminum can actually
uses about ninety five percent less energy than to make
(22:59):
that aluminum can from that virgin or And this gets
back to the point about the sort of cost spread
between recycled versus virgin feed stuff. Right, the harder and
more costly it is to make it virgin, the more
of a willing market there is for that recycled material.
Speaker 1 (23:16):
So cans are great, they always work as a business.
Aluminum cans are good. Okay, what's next.
Speaker 2 (23:20):
Aluminum cans are great. Next is we're going to get
a little technical here. HDPE natural. So this is HDPE
is a type of plastic resin. If you look at
the little Chasing Arrows recycling logo, it's resin number two
and most commonly takes the form of milk jugs. Right,
that's sort of translucent white.
Speaker 1 (23:38):
The gallon milk jug exactly exactly.
Speaker 2 (23:41):
Yeah, and then from there, you know, I'd say it's
probably pet bottles. So that's triangle number one. That's like
your water bottles, your soda bottles. This is actually a
type of resin where we forecast a huge gap in
the supply versus what's going to be demanded about five
years from now. So that's a really interesting interesting one
to watch.
Speaker 1 (24:01):
And then why I can imagine demand going up, But
why can't they just make more of them from virgin petroleum.
Speaker 2 (24:09):
Yeah. It's a combination of legislative requirements around minimums recycled
content combined with sort of like the nature of the
end markets that are demanding pet. So you know, pet
could be used by the you know, water beverage bottle manufacturers,
but a lot of that pet also gets absorbed into
(24:29):
markets you wouldn't imagine, like carpet or mattresses or other textiles.
Speaker 1 (24:34):
So the bad news is that there's plastic everywhere. But
the good news is at least they can use their
recycled bottles.
Speaker 2 (24:40):
Exactly exactly, so that that's one turf that's getting pretty heated.
And then to at least a round out the sort
of container side of things. The other very common thing
that gets sorted is HGPE color. So h again it's
triangle number two, but it's it's been dyed, right. So
this is typically things like your shampoo bottles or your
laundry detergent.
Speaker 1 (25:00):
Jumps, and so is that also like pretty good? Are
we still at pretty good on the stack? That's all.
Speaker 2 (25:05):
That's all pretty good. I'd say those every single recycling
facility you go to will sort out those commodities. Okay,
And I'll mention here that there's a big time honorable
mention for cardboard and for paper. I was focusing on containers,
but a lot of people talk about plastics a ton.
The majority of the recycling stream is still paper and cardboard, right,
so that stuff like that's almost table stakes for a
(25:27):
recycling facility. Just you have to get that right if
you want to stay profitable.
Speaker 1 (25:31):
And that's a reasonable business as well.
Speaker 2 (25:33):
Yes, that's a very reasonable business. A lot of these
recycling facilities actually talk about something called the Amazon effect.
In other words, as you know, e commerce and shipping
has become the way we buy things. Cardboard has just
inundated the recycling stream, which is great because you can
always sell cardboard now there's a really hot market, but
also not so great because maybe your facility was built
(25:55):
ten to twenty years ago before this became a thing,
and now you have to find a way to sort
of jerry rig it to handle all of these massive,
oversized boxes that are coming through your stream.
Speaker 1 (26:03):
So what that we recycle is dumb, isn't really getting recycled,
doesn't make sense whatever.
Speaker 2 (26:09):
Yeah, this is a very long tail of things.
Speaker 1 (26:11):
You know I mentioned for everything else. Yeah, it is.
Speaker 2 (26:14):
The everything else. And this points me to what I
often tell people is a misconception. But recycling is the
phenomenon of wish cycling and I was guilty of this
too until I started Glacier and learn a bit more,
which is this idea that if you're not sure if
something is recyclable, a lot of people who want to
do good for the planet are like, I'll toss it
into the recycling bin in case they can do something
with it. And in fact, this ends up being a
(26:35):
huge problem for these facilities because most of the time
they can't do something with it, much as we wish
they could. So a lot of these contaminants that people
throw in there are things like those plastic bags or
you know, those films and flexibles, which some facilities can recycle,
but most can't. It's things that have plastic in them,
but it's not kind of standard plastic. So for example,
(26:59):
one very confusing and insidious example that gets brought up
often is children's toys. Right, maybe they're made out of
some bulky plastic. You're like, hopefully this can be recycled,
but chances are that toy has various different grades of
plastic on it that aren't easy to pull apart. And
heaven forbid it's an electronic toy with the batteries still
in it, because that can literally cause an explosion or
(27:20):
a fire and blow up the facility.
Speaker 1 (27:26):
We'll be back in a minute with the lightning round.
Let's do a lightning round. It's going to be a
little more random. So I know you were a consultant.
Is it right that you were a management consultant?
Speaker 2 (27:45):
I was a management consultant.
Speaker 1 (27:47):
I'm curious you know now you run a company, right,
I'm curious what do you know now from running a
company that you wish you knew when you were you know,
telling people how to run their company.
Speaker 2 (28:00):
Yeah. I think a lot of what I've learned running
Glacier has been around always identifying and then sort of
like revalidating what is that north star metric or objective
that we're aiming for, and then making sure that anything
else we're working on or anything we're communicating is in
(28:21):
support of that.
Speaker 1 (28:22):
So, like, you're still a consultant.
Speaker 2 (28:26):
Absolutely, I mean, honestly, I think, like running Glacier, A
lot of people think that the tech is the hard part,
and don't get me wrong, it's insanely hard, insanely challenging.
But when you think about something as cross functional as
a circular economy, like I'd say, the majority of my
day gets spent thinking about how to align incentives right,
like recycling facilities, brands, manufacturers, local legislators, Like they all
(28:50):
kind of want different things, So how do you explain
initiatives or proposals to each of those parties in a
way that makes sense to them and gets everyone growing
in the same direction.
Speaker 1 (29:00):
So if the answer nothing is the answer, you feel
like you're actually still doing what you were doing.
Speaker 2 (29:04):
No, I mean I would say that, you know, one
big mindset shift for me that has been very healthy
is I'm definitely a perfectionist and a type a personality.
Speaker 1 (29:14):
You know, by.
Speaker 2 (29:14):
Upbringing and in management consulting, you're really encouraged to lean
into that, right, Like people are paying you big bucks
to make sure that you got every every single last
detail down to the decimal place right everywhere. And so
you know, I'd say that my consulting days were great
for training me on like how to make sure I
(29:35):
knew what details mattered and really like make sure that
everything lined up. But with an early stage startup, it's
the opposite where like you don't have time to be
perfecting everything, and so that has actually allowed me to
sort of flex towards how quickly can I move and
still be efficient, Right, Like, what is the right sort
of balance of making sure that you're putting out high
(29:59):
quality work and that things are generally moving the right direction,
but also realizing that actually it's okay and probably good
that certain balls are getting dropped, because, as one of
my mentors told me, if you find that you are
doing everything perfectly and nothing is failing, you're probably not
moving fast enough.
Speaker 1 (30:16):
Yeah. I interviewed the guy who started planning at the
satellite company, and he told me that he was upset
when none of their satellites were failing. It meant they
weren't launching soon enough, they were spending too long to
work on it. It's the same, That's exactly right.
Speaker 2 (30:30):
That's been a massive learning and frankly a pretty painful
one in the early years of Glacier, when all I
wanted was to make sure that every single thing I
outputed was going to work. And you know, at the
end of the day, I was like, I just got
to get get rid of some of those sort of
controlling tendencies if I really want this company to scale
at the rate that it needs to.
Speaker 1 (30:47):
What's one tip to stop being to type A when
you're running a startup.
Speaker 2 (30:53):
Honestly, I don't know if this is healthy, but my
approach was to kind of just like overwhelm myself.
Speaker 1 (30:58):
You give yourself too many things to do so that
you have to just pass them on before you're done
with them.
Speaker 2 (31:03):
Yeah, and I'd say it wasn't our intentional per se,
because it's certainly not a very pleasant ex experience to
go through. But I often joke that, you know, I
think starting an early stage company was maybe the only
thing that could have broken me of some of these
perfectionist habits, because I really had to go through sort
(31:24):
of the dark side of pulling all nighters, working myself
to the bone, realizing, you know, like what is this
all for, and having that sort of existential crisis moment
to say, Okay, I don't want to give up on Glacier,
and I know we've got an immense amount of potential
ahead of us, so I now need to fundamentally rethink
how I'm balancing this list of a thousand priorities if
(31:46):
I want to do it and still be around in
a successful leader years from now.
Speaker 1 (31:51):
Are you less of a perfectionist in your non work
life now than you used to be?
Speaker 2 (31:55):
Uh? Absolutely? It's like amazing what perspective gives.
Speaker 1 (31:59):
You on things.
Speaker 2 (32:00):
Just a lack of time, yes, yeah, yeah, some would
call it just a raw lack of time. I do
think that it really sort of for you to think
much bigger picture about what matters to you and make
sure that you're carving time out for that and then
just not sweating the details. And the amazing thing is
once you start doing that and you realize that the
(32:21):
world isn't going to end because you forgot to do
this thing or decided not to do that thing perfectly,
it gets easier and easier to do right. So that's
been really healthy for me.
Speaker 1 (32:36):
Rebecca who Trums, is the co founder and CEO of Glacier.
Please email us at problem at pushkin dot fm. We
are always looking for new guests for the show. Today's
show was produced by Trinamnino and Gabriel Hunter Chang. It
was edited by Alexander Garreton and engineered by Sarah Briguerra.
(32:56):
I'm Jacob Goldstein and we'll be back next week with
another episode of What's Your Problem
Pushkin. Imagine you're running a factory. You've got to put
out a consistent, high quality product that your customers will buy,
but you have no control over the raw materials that
come into your factory every day. One day, half of
what comes in is literal garbage. The next mixed in
(00:38):
with your usual inputs is some random lithium ion battery.
It's a fire hazard and also a surfboard weird. And
you have to deal with all this stuff and still
keep getting your product out the door. This is literally
the way the recycling business works. Recycling plants take in
(00:59):
a largely random, occasionally hazardous stream of stuff, a stream
of stuff that changes in a pretty unpredictable way from
day to day, from hour to hour, and then recycling
plants have to turn that random stream of inputs into
aluminum and plastic and cardboard that other companies will buy
and use to make new stuff. This is why my
(01:22):
guest today calls recycling the most demented form of manufacturing
on the planet, and it's why she and her colleagues
are trying to use technology to bring some order to
the recycling chaos. I'm Jacob Goldstein, and this is what's
your problem the show where I talk to people who
(01:44):
are trying to make technological progress. My guest today is
Rebecca who Troms. She's the co founder and CEO of
a company called Glacier. Rebecca's problem is this, how do
you use AI and robotics to make recycling a somewhat
less demented business. If Rebecca and her colleagues are successful,
they'll not only help recycling plants work better, they'll help
(02:06):
companies figure out how to recycle more of this that
they're sending out into the world in the first place.
Our conversation started with Rebecca talking about the moment around
five years ago when she and her co founder decided
to start the company.
Speaker 2 (02:21):
I was, you know, even at the time, obsessed with trash,
just obsessed with where does all of our stuff go?
And it's one of those I call it like a
Matrix moment or a red pill moment where once you
realize that you've never thought about where all of your
garbage goes after you put your bins on the curb,
you can't unsee that, right. And so this was also
(02:42):
around the time where there was a lot of change
happening in the recycling industry. So we're rewinding to roughly
twenty eighteen twenty nineteen. One cataclysmic shift for the industry
is that China, who had previously been the world's largest
buyer of recycled feedstock to make into new things. They
basically said very rapidly, you know what we're taking in
(03:03):
the world's recycling, but most of this is trash. People
are not sorting it well enough. We're getting a ton
of contamination, and we don't want to end up as
the planet's landfill or incinerator. So we're going to drastically
increase the bar on quality of what we're accepting. And
then that caused shockwaves throughout the globe and certainly for
the US, where suddenly recyclers for the first time in
(03:23):
a long time, were like, the game is not to
just crank through all this recycled material, bail it and
ship it overseas. We actually need to invest a lot
more in increasing the bar on quality and on purity rate.
And what's even more challenging is that a lot of
the backbone, historically and even to this day, is still
just people standing next to conveyor belts sifting through our
(03:46):
recycling in our trash and of course that's not only
very dangerous, it's not a very well compensated job. There's
a lot of hazards there. But also there's this massive
sort of labor shortage.
Speaker 1 (03:57):
Yes, it seems like a robot friendly moment, a robot
friendly environment. So like, what's your move, what's your first move?
Speaker 2 (04:03):
So my co founder had already at the time, he
wasn't my co founder, he was just a friend of
a friend. He was already pretty intent on this idea
that hey, you know, automation AI, all of these technologies
are so good now that for the first time, we
could feasibly rapidly commercialize a purpose built industrial robot, specifically
(04:24):
for recycling sortation.
Speaker 1 (04:26):
Huh.
Speaker 2 (04:26):
And it's not going to cost us massive amounts of capital,
and we can actually do it in a matter of
like a couple of years.
Speaker 1 (04:33):
So this idea that oh, now is the moment? Is
it computer vision? Like, what is the underlying technology that
made five years ago or whatever? The moment when oh,
we can do this in a way that hasn't been
done before.
Speaker 2 (04:45):
Yeah, Honestly, it's the confluence of a lot of things.
So I'll break it down into sort of the computer
vision piece and then also the hardware or the robotics piece.
When you think about where industrial automation has come from
even to this day, a lot of those technologies are
operating in really well defined, truly repetitive roat environments. So
(05:06):
think about a robot at a warehouse, and it's literally
just palatizing identical boxes over and over again. And so
to harken back to this idea of recycling plants as
being this extremely volatile manufacturing environment, even if you have
automation that's just sorting let's say aluminum cans, right, You're
talking about aluminum cans that your computer vision needs to
(05:28):
detect in infinite varieties, not just thinking about the wide
variety of cans that are on the market in all
of their colors and designs, but also the fact that
they show up not as pristine cans but crinkled in
various ways, stuffed into bags. Like, there's so much heterogeneity
that even just identifying that item on conveyor belt that,
(05:50):
by the way, has dozens of other types of items
on it is already a massive challenge that only recently
has been something that we can adapt to in a
cost effective way. And then you can layer onto that
the fact that now you're not only seeing those items
with this computer vision system, but you also need to
find a way to actually go and grab that material
(06:10):
and sort it into the right location. So when we
talk about advances in sort of off the shelf parts
that you used to design and make your robot, or
even like the gripping technology that's available, a lot of
that even a decade ago, would have required an immense
amount of r and D, with a much bigger team
and a much higher price tag to get to the
same point that we've gotten to after just a couple
(06:32):
of years.
Speaker 1 (06:33):
Tell me about the first one you built. Tell me
about building a prototype and putting it in the world.
Speaker 2 (06:37):
Man, So there are a lot of different stages to
our prototyping. The first prototype, if you want to go
way back, was when I hadn't even decided to start
this company with my co founder yet, but I did
tell him I would help him learn about the industry
see if there was some sort of a business to
be had, and we met in his kitchen in San
Francisco where he had a little corner set up. So
(06:59):
like very very simple. I think that it actually involved
a used yogurt tub as like this rotating wheel with
a piece of string tied around it, Like that's how
janky we were talking, but it worked and we were like, Okay,
there's something here. The first piece of equipment we actually
installed into a recycling facility also tells you a lot
about the constraints that these facilities are under. I was
(07:21):
kind of thinking we had to have this super built out, sophisticated,
polished thing that we've proven out to the nines in
the lab, and we actually called up a number of
recycling facility operators nearby and one of them was like,
you know what, when you got something, just like bring
it in here and try it out, because literally I
remember him saying, if your robot can pick one more
(07:43):
can than I would have gotten otherwise, Like it's already
worth it to me.
Speaker 1 (07:47):
Just try it. Was there a rat moment, tell me more,
did you see a rat?
Speaker 2 (07:54):
Literally?
Speaker 1 (07:54):
I was like, what is that not a metaphor not
a metaphor a rodent?
Speaker 2 (07:59):
I have seen many rats. Literally, Even on that first
install there was a moment where I was like literally
army crawling under a conveyor belt to fasten one of
the legs of the robot, and I came eye to
eye with a rat, who then, of course ground a
piece of food that was on the floor and then
scurried away. Right, there are many friendly critters running around
some of these facilities.
Speaker 1 (08:19):
Was there any part of you that kind of loved it?
Speaker 2 (08:22):
Oh, one hundred percent.
Speaker 1 (08:24):
So let's talk about where you are today, both on
a kind of micro level, like what your robot or
robots look like, and then also a little more macro
of like the scope of the business. Do you have
like one basic robot what to look like or you
got a bunch of them?
Speaker 2 (08:38):
Yeah, so we have one base model of robot. We
actually are already working with several dozen customers across the country.
But if you imagine any sort of conveyor belt in
an industrial facility, our robot think of it almost like
a table, right, So it's got four legs and it
kind of sits over that belt, and then the guts
of the robot or the mechanisms doing the picking are
(09:00):
kind of over the top of that conveyor system. So
you've got these arms that are going back and forth.
They can pick up something from the belt and then
they carry it off to one of those sides of
the belt. Where they actually drop it into the right location. Okay,
so that's the robot. And now one other thing we
haven't really talked about is this computer vision system.
Speaker 1 (09:19):
So that's yeah.
Speaker 2 (09:20):
Imagine basically like a camera with some lights to illuminate
the belt sitting on this little rig that's a little
bit upstream of the robot. So the material passes under
the camera, the camera has a second to process or
a should I say, a couple milliseconds to process what
it's seeing, and then not only does it tell the
robot you know, hey, there's a can coming in this
(09:40):
location picket and then put it into this other spot,
but what we're also finding is that that data as
a standalone is also able to massively advance these operator's
abilities to understand and optimize their facilities. So that's opened
up a whole new world of use cases.
Speaker 1 (09:58):
Because they weren't gathering data in that kind of way before.
They only sort of knew what was coming through in
a very gross macro way.
Speaker 2 (10:07):
Right, And this gets back to you know, recycling facility operators.
I have so much admiration for how they have gotten
really resourceful with trying to understand their operations. But you know,
to give you a sense of things. The state of
the art in the industry to this day is still
mostly manual audits. And when I say that, I mean
imagine taking a half ton of material off your line
and then literally having two to four people hands sort
(10:30):
and categorize and wigh each item to understand what's coming through,
and then assuming that that half ton is representative of
like the thousands of tons coming through your facility on
a yearly basis. I often tell the story of one
of our early data customers, this gentleman who runs a
recycling facility in California. When he met me, he was
telling me that he had mounted a goat pro camera
(10:52):
above his conveyor belt, the one that was basically supposed
to be all trash leaving the facility, but he knew
he was missing some good stuff, and he would spend
an hour a day after work going frame by frame
through some random snippet and manually tallying how many cans
and bottles were on that line, and then using that
to back into what he would try and change in
(11:12):
his operation the next day, and then he would check
that day to see if it changed anything. And so
imagine his delight when I told him, hey, actually we
can mount our own camera on there, and suddenly we'll
just give you access to a dashboard.
Speaker 1 (11:24):
A machine will literally count everything that goes.
Speaker 2 (11:27):
Literally in real time. So we're seeing that, you know,
the robot is this incredible foot in the door with
a lot of our facility partners, but that everyone's starting
to realize that, hey, actually this data can also help
us understand the entire world, not just the location where
the robot is sitting either.
Speaker 1 (11:43):
So at this point is it kind of a robot
business on the front, but really you're like a computer
vision data AI business.
Speaker 2 (11:52):
So a lot of customers come to us saying, you know,
I literally had a gentleman tell me a couple months
ago like I would never pay you to tell me
what I already know about my trash. And I'm like,
I'm not going to convince you that you don't already
know everything about your trash. But you know you want
a robot, Let's get you a robot. And that has
sense of all the conversation where we're just sort of
starting to show him this data and he's like, oh,
(12:12):
actually I didn't realize that this was the case. The
flip side is also true where someone's like, I don't
know if I need a robot yet, but I'm really interested
to see what I'm losing on the back end. We
install that camera and then suddenly, lo and behold that
data makes the case that, holy cow, I'm losing so
much stuff. I don't just need one robot. I maybe
need two or even three robots. Right, So it's kind
of this mutually reinforcing flywheel that's been really integral to
(12:33):
the success of our business so far.
Speaker 1 (12:36):
Tell me about your work with Amazon and with Colgate Palmolith.
What are you doing for them with them?
Speaker 2 (12:41):
Yeah, So to start, maybe just I'd love to explore
this idea of what is a circular economy because it's
a buzzword that gets thrown out a lot, but it's
really important to understand why Amazon and Coliate and their
peers matter here. Right now, we are living in mostly
a linear economy. In other words, someone makes a thing,
we consume a thing, and then we dispose of it. Right.
(13:02):
A circular economy tries to turn that process into a circle.
So instead of throwing it out in a landfill forever
or incinerating it, that material gets brought back to the
front end and reused somehow to make new stuff that
we can then consume. And the ideal is to make
this go on forever so that we limit our resource consumption.
So now that we have this growing base of recycling
facilities that are you know, gathering data, that are getting
(13:25):
a better understanding of what's coming into their facilities, what's
actually being bailed and sent out to the end markets,
We're working with companies like Amazon and Colgate on a
number of fronts. You know, the first is even just
to understand where is all of that packaging going to
their credit They and several of their peers have realized
(13:45):
that there's a paradigm shift possible now from we have
designed a thing that's technically recyclable, you know, our packaging
R and D engineers have made this awesome mono material
HDPE toothpaste tube in Colgate's example, to now trying to understand, okay,
well we made this thing that is recyclable, is it
actually getting recycled? And that was a lens that we
(14:08):
couldn't really get at scale before and so now with
Glacier's technology, we're able to monitor in real time, you know,
how much of these tubes are actually ending up at
the recycling facility, and once they're in the facility, are
they ending up in the right place, are they being
sorted correctly such that they can actually be turned into
new stuff, or are they ending up in the landfill,
(14:29):
in which case you know, suddenly this recyclable tube isn't
very recyclable at all. So we're starting to answer these
really really critical questions.
Speaker 1 (14:37):
So Colgate knows whatever how many tubes of toothpaste they
sold in a city, and if you are working in
the recycling facility for that city, you can actually count
how many tubes of that toothpaste came down the recycling
line and how many tubes of that toothpaste wound up
in the bin. I mean, is that the reductive version
of what you're.
Speaker 2 (14:54):
Saying, essentially? Yeah, And where we're even seeing now, like
with these rapid advances in AI and in detection, that
first of all, it's no small feet to even define
what is a tube and how do you tell the
difference between a toothpaste tube versus a sunscreen versus a
lotion tube. But now we're getting to the point where
we can actually say the brand of toothpaste, it is
(15:16):
like from all of those visual markings, and so we're
just seeing this sort of Cambrian explosion of interest from
a wide variety of different you know, packaging producers and
brands to really start understanding this previous black box on
what happens once they release this packaging into the wild
for consumers to buy.
Speaker 1 (15:34):
So, I understand that California has a law that is
in some fashion supposed to put companies like on the
hook for their for their products right after they're used,
to incentivize companies to have their products be recycled. Right.
Is that am I characterizing that law right? And is
it relevant to your business?
Speaker 2 (15:53):
Yes? So I believe you're referring to EPR or extended
producer Responsibility laws. For those who may not have heard
of EPR, it's essentially this premise that, you know, if
our recycling and waste system is supposed to find a
way to do something good with all the stuff throwing out,
the people making all the stuff that we're throwing out
should probably have some skin in the game to make
(16:14):
sure that that stuff gets either disposed of or reused properly. Right,
And so EPR laws are already in effect throughout Europe,
throughout Canada, some other regions, and then they've been passed
in a number of states in the US, including California.
Now while EPR in the US is still in its infancy.
In other words, it's been passed in a number of states,
(16:36):
but there's a lot of hairiness to figuring out how
to actually implement the system across all the producers selling
into a state and all of the recyclers operating in
that state. It is, I think a step in the
right direction because in a lot of ways it helps
to create that circle we were talking about earlier. You know,
you're seeing that a lot of brands and producers are
(16:58):
starting to take even more of a vested interest in
understanding what is happening to all of their packaging, because
they know that imminently they're going to need to start
proving the sort of end of life outcomes for that
packaging in order to you know, one, not be heavily
fined and then two maybe even have a right to
continue selling into that state.
Speaker 1 (17:17):
It seems good that Amazon and Kolgay Palmolive are trying
to figure out if the things they make that are
recyclable are actually being recycled, but it seems like for
that sort of thing to happen at a meaningful scale,
you would need laws basically, right. I mean, if the
companies are just incurring the cost either out of the
goodness of their heart or in the hopes of you know,
generating goodwill that will lead to higher revenues, those seem
(17:41):
like marginal cases. Are the EPR laws such that you
think it will become a meaningful part of your business,
a meaningful part of the world, that companies will in
fact be on the hook to figure it out, or like,
what do you think is going to happen?
Speaker 2 (17:53):
You know, I will say that early indicators are that
all of these states are taking it quite seriously. So
in addition to requiring a lot of these brands and
producers to pay into a massive fund upfront to even
just start implementing some of this movement, a lot of
these states are also you know, we're seeing that some
of the kind of like early deadlines and fines for
(18:15):
non compliance are actually being upheld, which I think is
a really strong signal to the market. Hey, this is
something that needs to get taken seriously. Now to your point,
I do think that at the end of the day,
whatever flavor this legislation takes. The key to make sure
that recycling is still a viable and sustainable value proposition
(18:37):
is that there needs to be some sort of an
end market for that material, right because let's say these brands,
even if they're required to pay billions of dollars into
this EPR system, if there's no one on the back
end to receive that material that these recycling facilities are sorting,
then recycling can't really happen. At the end of the day.
Speaker 1 (18:57):
Someone needs to buy the bail of.
Speaker 2 (19:00):
Plastic exactly exactly. But if they can have guarantee that
there is a buyer on the other side, right that
that person or that company will buy at a certain price,
then suddenly they can sustain that business quite well for
the long run. And so to that point, you know,
one other model that's often brought up in the realm
(19:20):
of legislation is actually minimum recycled content laws, because it
kind of gets at the same issue from the other side,
where you say.
Speaker 1 (19:28):
Basically creating demand, creating demand for a bale of recycled
plastic coming out of the recycling facility.
Speaker 2 (19:35):
Exactly, and it kind of disentangles the market for recycled
feedstock from the market for virgin feedstock, which is another
great way to kind of catalyze the movement of that
material throughout that recycling ecosystem.
Speaker 1 (19:52):
We'll be back in just a minute. What are you
trying to figure out right now? What's a big thing
you're trying to figure out?
Speaker 2 (20:07):
We are at a really exciting inflection point, a glacier
because of I think two big things here. The first
is just how do we scale smoothly and rapidly. You know,
we've gone from a year ago we were making maybe
one robot every three months, and now we have the
capacity to make three to four robots per month, and
(20:28):
we're expecting to go even faster by the end of
this year in the next six months. And then the
other big frontier for us, in addition to just you know,
how do we scale and get more of our stuff
out there, is what the heck do we do with
all of this data? Right? We've already seen that the
early use cases for this information people are taking to
really in droves. But there's a much more built out
(20:51):
version of the data platform where we say, you know,
we don't just have a camera in one or two
points throughout your facility. We can actually get sensors to
sort of blanket the facility, and in that regard, we
can take a huge step towards becoming more like that
manufacturer who has information on every single step of their
process and respond in real time and know exactly what's
(21:11):
going on at each stage.
Speaker 1 (21:13):
So, I mean, let's just talk about the recycling business
for a minute. The facilities you're talking about, they're just
private companies.
Speaker 2 (21:18):
The vast majority of them are. So this is a
common misconception about recycling, is that you know, these are
all you know somehow like public entities buy our estimation
about eighty or eighty five percent of these recycling facilities
are privately owned, and they can be anything from a
family owned business all the way up to the massive
waste companies like Waste Management, Republic Services, Waste Connections. These
(21:41):
are publicly traded companies that also own many of these
recycling plants as well.
Speaker 1 (21:46):
And the recycling plants are buying recycling from municipalities like
they do they pay for whatever tans and plastic jugs.
Speaker 2 (21:57):
It's actually a very interesting question. It depends a lot
on the condition of those end markets. We talked about
so in today's climate where those markets are really volatile
and a little bit uncertain, oftentimes, you know, these recycling
facilities will get paid by municipalities in order to take
any process that material. But what's interesting is, you know,
(22:17):
back during the heyday of recycling, when you know, China
was buying everything, there was no shortage of you know,
appetite for that material. The equation flipped.
Speaker 1 (22:27):
My sense is some recycling is quite efficient and a
good business, and some is not very efficient in a
bad business. Right, give me the like stack ranking for
recycling best thing to recycle to worst.
Speaker 2 (22:38):
Yeah, the I mean, at the end of the day,
the best things to recycle, according to a recycling facility operator,
would be the things that most reliably will make you
the most money. So top of the stack would be
aluminum cans, because there's always a market for those. They're
super easily recyclable, and to recycle an aluminum can actually
uses about ninety five percent less energy than to make
(22:59):
that aluminum can from that virgin or And this gets
back to the point about the sort of cost spread
between recycled versus virgin feed stuff. Right, the harder and
more costly it is to make it virgin, the more
of a willing market there is for that recycled material.
Speaker 1 (23:16):
So cans are great, they always work as a business.
Aluminum cans are good. Okay, what's next.
Speaker 2 (23:20):
Aluminum cans are great. Next is we're going to get
a little technical here. HDPE natural. So this is HDPE
is a type of plastic resin. If you look at
the little Chasing Arrows recycling logo, it's resin number two
and most commonly takes the form of milk jugs. Right,
that's sort of translucent white.
Speaker 1 (23:38):
The gallon milk jug exactly exactly.
Speaker 2 (23:41):
Yeah, and then from there, you know, I'd say it's
probably pet bottles. So that's triangle number one. That's like
your water bottles, your soda bottles. This is actually a
type of resin where we forecast a huge gap in
the supply versus what's going to be demanded about five
years from now. So that's a really interesting interesting one
to watch.
Speaker 1 (24:01):
And then why I can imagine demand going up, But
why can't they just make more of them from virgin petroleum.
Speaker 2 (24:09):
Yeah. It's a combination of legislative requirements around minimums recycled
content combined with sort of like the nature of the
end markets that are demanding pet. So you know, pet
could be used by the you know, water beverage bottle manufacturers,
but a lot of that pet also gets absorbed into
(24:29):
markets you wouldn't imagine, like carpet or mattresses or other textiles.
Speaker 1 (24:34):
So the bad news is that there's plastic everywhere. But
the good news is at least they can use their
recycled bottles.
Speaker 2 (24:40):
Exactly exactly, so that that's one turf that's getting pretty heated.
And then to at least a round out the sort
of container side of things. The other very common thing
that gets sorted is HGPE color. So h again it's
triangle number two, but it's it's been dyed, right. So
this is typically things like your shampoo bottles or your
laundry detergent.
Speaker 1 (25:00):
Jumps, and so is that also like pretty good? Are
we still at pretty good on the stack? That's all.
Speaker 2 (25:05):
That's all pretty good. I'd say those every single recycling
facility you go to will sort out those commodities. Okay,
And I'll mention here that there's a big time honorable
mention for cardboard and for paper. I was focusing on containers,
but a lot of people talk about plastics a ton.
The majority of the recycling stream is still paper and cardboard, right,
so that stuff like that's almost table stakes for a
(25:27):
recycling facility. Just you have to get that right if
you want to stay profitable.
Speaker 1 (25:31):
And that's a reasonable business as well.
Speaker 2 (25:33):
Yes, that's a very reasonable business. A lot of these
recycling facilities actually talk about something called the Amazon effect.
In other words, as you know, e commerce and shipping
has become the way we buy things. Cardboard has just
inundated the recycling stream, which is great because you can
always sell cardboard now there's a really hot market, but
also not so great because maybe your facility was built
(25:55):
ten to twenty years ago before this became a thing,
and now you have to find a way to sort
of jerry rig it to handle all of these massive,
oversized boxes that are coming through your stream.
Speaker 1 (26:03):
So what that we recycle is dumb, isn't really getting recycled,
doesn't make sense whatever.
Speaker 2 (26:09):
Yeah, this is a very long tail of things.
Speaker 1 (26:11):
You know I mentioned for everything else. Yeah, it is.
Speaker 2 (26:14):
The everything else. And this points me to what I
often tell people is a misconception. But recycling is the
phenomenon of wish cycling and I was guilty of this
too until I started Glacier and learn a bit more,
which is this idea that if you're not sure if
something is recyclable, a lot of people who want to
do good for the planet are like, I'll toss it
into the recycling bin in case they can do something
with it. And in fact, this ends up being a
(26:35):
huge problem for these facilities because most of the time
they can't do something with it, much as we wish
they could. So a lot of these contaminants that people
throw in there are things like those plastic bags or
you know, those films and flexibles, which some facilities can recycle,
but most can't. It's things that have plastic in them,
but it's not kind of standard plastic. So for example,
(26:59):
one very confusing and insidious example that gets brought up
often is children's toys. Right, maybe they're made out of
some bulky plastic. You're like, hopefully this can be recycled,
but chances are that toy has various different grades of
plastic on it that aren't easy to pull apart. And
heaven forbid it's an electronic toy with the batteries still
in it, because that can literally cause an explosion or
(27:20):
a fire and blow up the facility.
Speaker 1 (27:26):
We'll be back in a minute with the lightning round.
Let's do a lightning round. It's going to be a
little more random. So I know you were a consultant.
Is it right that you were a management consultant?
Speaker 2 (27:45):
I was a management consultant.
Speaker 1 (27:47):
I'm curious you know now you run a company, right,
I'm curious what do you know now from running a
company that you wish you knew when you were you know,
telling people how to run their company.
Speaker 2 (28:00):
Yeah. I think a lot of what I've learned running
Glacier has been around always identifying and then sort of
like revalidating what is that north star metric or objective
that we're aiming for, and then making sure that anything
else we're working on or anything we're communicating is in
(28:21):
support of that.
Speaker 1 (28:22):
So, like, you're still a consultant.
Speaker 2 (28:26):
Absolutely, I mean, honestly, I think, like running Glacier, A
lot of people think that the tech is the hard part,
and don't get me wrong, it's insanely hard, insanely challenging.
But when you think about something as cross functional as
a circular economy, like I'd say, the majority of my
day gets spent thinking about how to align incentives right,
like recycling facilities, brands, manufacturers, local legislators, Like they all
(28:50):
kind of want different things, So how do you explain
initiatives or proposals to each of those parties in a
way that makes sense to them and gets everyone growing
in the same direction.
Speaker 1 (29:00):
So if the answer nothing is the answer, you feel
like you're actually still doing what you were doing.
Speaker 2 (29:04):
No, I mean I would say that, you know, one
big mindset shift for me that has been very healthy
is I'm definitely a perfectionist and a type a personality.
Speaker 1 (29:14):
You know, by.
Speaker 2 (29:14):
Upbringing and in management consulting, you're really encouraged to lean
into that, right, Like people are paying you big bucks
to make sure that you got every every single last
detail down to the decimal place right everywhere. And so
you know, I'd say that my consulting days were great
for training me on like how to make sure I
(29:35):
knew what details mattered and really like make sure that
everything lined up. But with an early stage startup, it's
the opposite where like you don't have time to be
perfecting everything, and so that has actually allowed me to
sort of flex towards how quickly can I move and
still be efficient, Right, Like, what is the right sort
of balance of making sure that you're putting out high
(29:59):
quality work and that things are generally moving the right direction,
but also realizing that actually it's okay and probably good
that certain balls are getting dropped, because, as one of
my mentors told me, if you find that you are
doing everything perfectly and nothing is failing, you're probably not
moving fast enough.
Speaker 1 (30:16):
Yeah. I interviewed the guy who started planning at the
satellite company, and he told me that he was upset
when none of their satellites were failing. It meant they
weren't launching soon enough, they were spending too long to
work on it. It's the same, That's exactly right.
Speaker 2 (30:30):
That's been a massive learning and frankly a pretty painful
one in the early years of Glacier, when all I
wanted was to make sure that every single thing I
outputed was going to work. And you know, at the
end of the day, I was like, I just got
to get get rid of some of those sort of
controlling tendencies if I really want this company to scale
at the rate that it needs to.
Speaker 1 (30:47):
What's one tip to stop being to type A when
you're running a startup.
Speaker 2 (30:53):
Honestly, I don't know if this is healthy, but my
approach was to kind of just like overwhelm myself.
Speaker 1 (30:58):
You give yourself too many things to do so that
you have to just pass them on before you're done
with them.
Speaker 2 (31:03):
Yeah, and I'd say it wasn't our intentional per se,
because it's certainly not a very pleasant ex experience to
go through. But I often joke that, you know, I
think starting an early stage company was maybe the only
thing that could have broken me of some of these
perfectionist habits, because I really had to go through sort
(31:24):
of the dark side of pulling all nighters, working myself
to the bone, realizing, you know, like what is this
all for, and having that sort of existential crisis moment
to say, Okay, I don't want to give up on Glacier,
and I know we've got an immense amount of potential
ahead of us, so I now need to fundamentally rethink
how I'm balancing this list of a thousand priorities if
(31:46):
I want to do it and still be around in
a successful leader years from now.
Speaker 1 (31:51):
Are you less of a perfectionist in your non work
life now than you used to be?
Speaker 2 (31:55):
Uh? Absolutely? It's like amazing what perspective gives.
Speaker 1 (31:59):
You on things.
Speaker 2 (32:00):
Just a lack of time, yes, yeah, yeah, some would
call it just a raw lack of time. I do
think that it really sort of for you to think
much bigger picture about what matters to you and make
sure that you're carving time out for that and then
just not sweating the details. And the amazing thing is
once you start doing that and you realize that the
(32:21):
world isn't going to end because you forgot to do
this thing or decided not to do that thing perfectly,
it gets easier and easier to do right. So that's
been really healthy for me.
Speaker 1 (32:36):
Rebecca who Trums, is the co founder and CEO of Glacier.
Please email us at problem at pushkin dot fm. We
are always looking for new guests for the show. Today's
show was produced by Trinamnino and Gabriel Hunter Chang. It
was edited by Alexander Garreton and engineered by Sarah Briguerra.
(32:56):
I'm Jacob Goldstein and we'll be back next week with
another episode of What's Your Problem
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
The Breakfast Club
The World's Most Dangerous Morning Show, The Breakfast Club, With DJ Envy And Charlamagne Tha God!
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.