Latest / The ReForge Podcast / Five Thousand Robots a Year to Break Even (ReForge 032)

Transcript

Five Thousand Robots a Year to Break Even (ReForge 032) · The ReForge Podcast

• 0:05
  Welcome to the Reforge podcast.
• 0:07
  I'm Matthew Chang, and we're here to explore all things AI, automation, and the American worker, proving that the American re-industrialization is possible, pivotal, and pressing.
• 0:19
  Thanks for joining us today.
• 0:21
  Let's get into the episode Welcome everyone to episode 32 of the Reforged Podcast.
• 0:30
  We are still rocking and rolling, and we have got some great content for you today.
• 0:37
  Today, we're gonna talk about one of our favorite topics, Matt's a little iffy about them.
• 0:42
  But we're gonna talk robots.
• 0:44
  How's that sound, Matt?
• 0:45
  Yeah.
• 0:46
  Yeah.
• 0:46
  I'm very uncomfortable with this topic.
• 0:48
  I'm out of my depth.
• 0:49
  Yeah, I mean, the Chang Robotics name itself is a real harbinger for that, but, you know, we'll, we'll get past it.
• 0:56
  Yeah.
• 0:56
  All righty, so let's jump right into it.
• 0:59
  Um, so recently, um, uh, the Agility Robotics, uh, they came out with- I, I guess drawing a line in the sand if you want to.
• 1:13
  They, they kind of put it out there, um, all in on...
• 1:17
  Really, it seems like all in on the robot as a service model.
• 1:21
  Uh, we've talked about this a little bit.
• 1:23
  Um, you know, I, I tend to agree with the idea that I think the outright buying of it is the appropriate, uh, solution just from a, a financial engineering perspective.
• 1:36
  Um, but it, it looks like they are really kind of going all in or, or really trying to lean towards the, the, the robot as a service.
• 1:48
  I mean, that's what they, they seem to want.
• 1:51
  What...
• 1:52
  Let, let's get your...
• 1:53
  I mean, you, you had some strong comments on it.
• 1:55
  Um, not just that he's cute, um, but- Well, it is cute I mean, let's talk about it.
• 2:01
  It is.
• 2:01
  I wanna go on the record and say it's cute.
• 2:03
  I just hope that those eyes aren't color changing,'cause when you see them go red that's gonna be kind of terrifying.
• 2:09
  Well, there's a lot of safety lighting in automation.
• 2:12
  Maybe red means we're having a malfunction and come, you know, check my battery or something.
• 2:16
  Who knows?
• 2:17
  But yeah, so d- I think you did some research.
• 2:21
  What, what was the trailing revenue on Agility?
• 2:23
  So 1.5 million.
• 2:27
  My, my understanding is that they've been around for a while, like 10 years, and they spun out of a university research lab.
• 2:34
  Does that sound right?
• 2:36
  Yep.
• 2:37
  Yeah.
• 2:38
  And so, you know, it, uh, no doubt started off as a bunch of smart guys that were trying to make robots do novel things.
• 2:46
  I think, check, they have accomplished that.
• 2:48
  Um, I've watched some videos of Agility Robotics for more than a year, and I've thought to myself,"That's pretty nifty.
• 2:57
  I wouldn't mind grabbing a few of those and doing something specifically at, like, end of line in a manufacturing center- Yeah where I have, like, cases or boxes, something predictable coming out, and then Agility can grab the parcel and then go sort it or deposit it somewhere.
• 3:14
  Um, one point of differentiation, I think this is being branded on Wall Street as, uh, you know, a humanoid company, but to me, I...
• 3:25
  When I looked at it, I was like,"It's a kangaroo company." Right.
• 3:28
  If you look at the- The legs.
• 3:29
  So that, that was, that was the first thing I thought of when I was like, I was like, wait a second.
• 3:32
  And the, the videos are wild of them.
• 3:34
  Like, it's just, it doesn't seem- Well, assuming the- It's just wild assuming the videos haven't been sped up, the, the movement and the rate is actually pretty impressive.
• 3:43
  They get around pretty good.
• 3:45
  So I wanna start with a compliment to that team.
• 3:47
  Like, they built a, a killer robot.
• 3:50
  Uh, I like the novel form factor.
• 3:53
  I like that they went with the kangaroo design instead of the human.
• 3:57
  And it calls into question some of this bipedal, uh, operation that you're seeing from the, the more trendy humanoid companies.
• 4:04
  But, you know, leaving that aside, the, it, it, the economics, it's very clear that the original inventors of the company are no longer in control.
• 4:15
  That's, that should jump right off the page at you, because what you have here is a probably well-intentioned but poorly executed pricing model.
• 4:26
  So to understand what would drive a large company that has a lot of logistics to wanna go with the Agility Robot, it's a, it's a labor challenge, right?
• 4:36
  This is a pure cost savings play.
• 4:39
  This robot- Yep is designed as a retrofit.
• 4:41
  It goes into an existing facility.
• 4:43
  Most, its most, uh, popular task is ca- probably gonna be what's called zone pick in a e-commerce facility.
• 4:51
  So zone pick is where you have a bunch of aisleways of items, and human shoppers go around in the zone, and they pick the items that are ordered.
• 4:58
  So one tube of toothpaste, one toothbrush, a pack of crayons.
• 5:02
  You can tell I have little kids based on what I'm ordering here.
• 5:05
  And, um, and then they throw- Two cents they throw it in a box or a tote, and they send it downstream to a pack-out station, where it's gonna get boxed, labeled, and, and then sorted for ship, right?
• 5:17
  So the, the Agility Robot form factor is really well suited for a zone pick that assumes no retrofit.
• 5:24
  Uh, but when we dig into the economics, you know, in my, uh, my post there, I had, uh, some, just a quick calc on it.
• 5:31
  To make a dent in, in the operation, you're gonna need a couple dozen of these things.
• 5:36
  So on, on the magnitude of, say, three dozen, 36 robots.
• 5:41
  Uh, the challenge is the five-year cost is anywhere from, from 14 to 18 million, depending on which, if you go on the rent column or the buy column.
• 5:51
  And at 14 to 18 million to run a zone of zone pick, just a, you know, some shopping aisles, that is tough, man.
• 6:00
  That, uh, that is really tough.
• 6:02
  I mean, 8,500, if you assume, you know, uh, all right, let's say, oh, you're, you're kind of fully burdened employee, you know, 60% of that, you know, we'll say comes to salary.
• 6:16
  You know, you're talking about you, you've, you're paying 61K a year for, for an employee, you know, working full time It would appear this is suited for an environment where you're not operating 40 hours a week per, per shift.
• 6:40
  The, the value comes from being able to take that up to multiples of it, because at 61K a year, you know, for sa- offering 61K for a salary plus benefits on top of that, you're...
• 6:56
  I, I'm struggling to see the, the, the real economic value there for the use cases that y- you mentioned.
• 7:03
  I mean, it's- Yeah, I th- I think I threw in a s- a 16% internal rate of return.
• 7:08
  So again, for the audience, uh, Colin and I are taking a finance bias here on this show today.
• 7:14
  We're not doing technical.
• 7:16
  We already, we gave the technical comments up front.
• 7:18
  Um, the, uh, this is a purely financial decision for any business.
• 7:24
  If it's a cost saving initiative, we just have to evaluate is what we're saving enough to offset the costs of what we're paying.
• 7:31
  Um, so a couple, couple underlying conditions.
• 7:34
  Number one, labor for box movers and pickers is getting tough, man.
• 7:39
  That is directly impacted by increase in trades.
• 7:44
  It's dec- impacted by the change in, uh, demographics, and it's also, um, being hugely impacted by deportations, right?
• 7:52
  So a lot of that labor that is gonna do each's pick or zone pick or box moving is, um, is, you know, non-English speaking, non-citizen labor in the United States.
• 8:04
  So they are in the sweet spot for the change in demographics.
• 8:08
  But in terms of the pricing, I mean, for that robot to spend 400,000 or 500,000 is insane.
• 8:15
  That's absolutely insane.
• 8:16
  There's no way it's ever gonna make the productivity to pay back on that.
• 8:19
  So when I said, calculated a 16% IRR, uh, that's the primary metric that a CFO or a VP of finance at any corporation is gonna use to cost justify a cost savings.
• 8:31
  So it equates to a six and a half, 6.25 year payback.
• 8:36
  That ain't gonna make it in e-commerce.
• 8:38
  And, and if you think about most e-commerce, it's, um, like FBO by Amazon or something.
• 8:44
  It's either consignment inventory where you're holding another company's inventory and you're doing the fulfillment to the e-commerce, um, or it's a, a 3PL operation, which are typically signed, or 4PL, signed up on a three to five year contract for distribution.
• 9:01
  So when the payback window is outside the customer contract period, it's kind of a non-starter.
• 9:07
  You're dead on arrival.
• 9:08
  Um, also there's not...
• 9:11
  Yeah, assume a fair gross margin.
• 9:13
  You know, let's give them a 25% gross margin.
• 9:16
  There's not 300K worth of parts in that robot, right?
• 9:19
  So- No so what, what you see here is, um- Is primarily designed for investors is to say that we have an enormous backlog and that because there probably are no doubt a lot of companies that express interest and signed a letter of intent to get access to the first deployment of Agility robots.
• 9:39
  Um, the problem is none of that capital has been appropriated for the purchase.
• 9:45
  It's just a LOI for kind of a pilot or for a marketing, uh, stunt.
• 9:51
  And that's what you see companies like Figure doing, is they're doing pilots and marketing.
• 9:55
  They're not actually used in operations and production.
• 9:58
  So that's gonna be the gap here, is you got this huge, um, gross margin or contribution margin sitting on top of what's probably 80K worth of parts.
• 10:10
  You know, the, uh, the final cost of assembly of that robot, if I had to guess, is probably in the, you know, as low as 50,000, but probably in the$80,000 range if it's...
• 10:18
  I don't know where it's made.
• 10:20
  Um- I mean, because we, we know that it's not domestically, you know, all sole sourced from domestic components.
• 10:25
  There's, the supply chain isn't there.
• 10:27
  So they're, they're definitely enjoying a nice, nice margin on that Yeah, so well, and, and those import- imported parts are now subject to controls and, uh, it has, it is gonna have to be assembled here because of the latest, um, FTC bans on imported autonomous robots.
• 10:44
  That, that's, that's what I was, uh, gonna get to, is I, I think that there's changes that are taking place right now where, uh, to some extent it seems like maybe they're trying to be an early mover, early to really get these out there and offer it- Totally, and they did which I get.
• 11:00
  Success.
• 11:00
  Yeah.
• 11:01
  Yeah, yeah.
• 11:01
  I mean, but so I, I looked it up.
• 11:03
  They did, um, 1.8 million, uh, gross.
• 11:06
  They, uh, and it was, so 1.8 million in net sales they say according to the S4 for their, uh, upcoming SPAC.
• 11:16
  Um, 140 million operating loss, 111 million on o- operations, up from 71 million in 2024, and they burnt about 100 million in cash.
• 11:30
  So- Ooh they're looking- Those are tough, tough numbers, man.
• 11:34
  Tough They're doing a, a$2.5 billion SPAC looking to generate 620 million in gross proceeds.
• 11:44
  Um, so a- a- this is to me looks like something that is intended to be, you know, part of the hype ahead of that SPAC, which- Yeah uh, under- understand, certainly understand those mechanics, but I can't help but have this question in the back of my head of what are we gonna see in the next 18 to 24 months i- in terms of parallel or, or competitive offerings in this space?
• 12:12
  Where does that price point come in, in ter- i- in terms of being competitive'cause- Yeah, I mean, the, the low model on the Unitree, if you're just gonna buy the Unitree humanoid outright, the low model is one-month rental on the Agility, right?
• 12:27
  Now it, it, Agility, let's argue that it has a more full stack.
• 12:30
  It's got the opportunity to charge, which their charger, charging system is sick, if you've seen the video.
• 12:36
  Robot just goes over to a wall mount.
• 12:38
  It's kind of like a Dyson vacuum cleaner, how it's got a little wall mount, and it hangs itself up, and it kind of goes to sleep while it charges.
• 12:44
  Um, so again- It's gonna look like a rollercoaster going through the factory super cute.
• 12:49
  Super cute, super efficient, but what, what the robot's not built for is long haul, so it's confined to a workspace.
• 12:55
  The, the most likely workspace is the zone pick of an e-commerce fulfillment center or a 3PL sort facility.
• 13:02
  Um, the economics, um, I mean, they're just never gonna make sense.
• 13:06
  So what, so what you have is this pricing sheet was produced for investors, and the investors no doubt know that the, the finished good cogs on the robot inclusive of parts, tariffs, assembly, the whole thing- Probably about 80,000 is my guess.
• 13:25
  So then the rest of it is pure margin.
• 13:28
  So what, what you're trying to get to here for investors is a SaaS margin on hardware, right?
• 13:33
  That's what the advertising is.
• 13:34
  Well, and they have the annual subscriptions that go with it too, for the actual, you know, the, the software maintenance of 36 pay- So then you take that, take, and you say it's at a six, six and a half year ROI.
• 13:48
  I did$18 an hour fully loaded, 80% utilization, 350 days a year to get that calculation, and a 10% weighted average cost of capital, which is about industry standard.
• 14:00
  So I took, you know, kind of down the fairway assumptions to get there.
• 14:04
  Um, applied a burden rate to the labor, you know, for workers' comp and all that.
• 14:08
  Um, and, uh, but you're replacing$18 labor here.
• 14:12
  Maybe, maybe$22 labor in a premium market like Southern California, but this is low cost labor that you're replacing with Agility.
• 14:20
  Remember, it doesn't have dexterity, so it doesn't have the hand that can go into the bin and pick out the item- It's point A to point B and package the box, right.
• 14:28
  It's just moving the box.
• 14:30
  That's all it's doing, the bin or the box.
• 14:32
  So you're replacing, you know, v- obviously the lowest end of the wage scale.
• 14:37
  You're at a six plus year payback.
• 14:39
  Agility, to get a big order, let's say they've got interest from Walmart, let's say they've got interest from Target or Amazon or any big, uh, box mover, then, um, those guys are...
• 14:52
  I mean, they're no dummies in procurement.
• 14:54
  They've been buying technology forever.
• 14:56
  So they're gonna require a discount to get it within the payback window.
• 15:01
  At this point, with$140 million spend to get$1.5 million revenue, Agility is buying customers, right?
• 15:08
  Oh, yeah.
• 15:08
  They're subsidizing customers.
• 15:10
  At that pricing sheet, that means they only have 18 robots deployed at 1.8 million trailing revenue.
• 15:16
  So, so it's a relatively small thing.
• 15:18
  They have to scale up manufacturing here and they have to do the, what, what do they call it on, on X?
• 15:22
  The, um, forward deployed engineer, the FDE.
• 15:25
  Oh, yeah.
• 15:25
  The, the, the FDE.
• 15:27
  Which we, which you and I just call engineer, right?
• 15:30
  So they have the deployment engineers and the system architects,'cause one of, one of our, uh, mutuals, uh, Mike Cobbler, he pointed out, uh, who's gonna do the ERP integration?
• 15:43
  Accenture's gonna love this.
• 15:44
  Capgemini, Colin, is gonna love this, right?
• 15:47
  PWC will love it, Deloitte, all of them.
• 15:50
  I mean- Yeah it's, that, that was one of the thoughts I had is, okay, so you, you've added a whole other level of data coming in streams that not only there's the, obviously the whole cybersecurity element of it to not get into, but you've now got work orders or, or basically worksheets that you're dispatching dynamically to these.
• 16:12
  Uh, that first few years of this, it's going to be a learning experience because- That's right I would bet money that these, the, whatever the operating system is for these robots, it is not designed for kind of that ERP integration first, which, you know, I'm sure it's a thought in there, but why they're, they're not designing it for ERP integration, they're designing it to do the robot thing And so- That's right it, it's just, it's, it- Okay, you pulled me into technical you're paying a lot more for the deployment.
• 16:44
  I, I was just gonna do the, the business economics, but you pulled me into technical.
• 16:48
  So for the audience- But that is, it's part of it.
• 16:50
  Think about, you know, all those costs- Totally you've got.
• 16:52
  You have to capitalize all of that, th- those costs for that new capital project you've brought in there.
• 16:56
  You've got everything that goes into it.
• 16:58
  You've just added, you know, 5 million- Oh, yeah.
• 17:01
  Oh, yeah is a very cheap, I would say probably 10 times that if you actually do a full-scale fleet deployment of these.
• 17:08
  Yeah.
• 17:08
  So, so what, what happens in all these facilities, they have a WMS, warehouse management system, or they have a ERP.
• 17:16
  And, uh, and those software systems that are kind of the, those are the enterprise software systems and inventory systems of record.
• 17:23
  They're gonna receive an order from a website or from a- another store, like a CVS might, might wanna use this for in-store replenishment.
• 17:32
  So they receive an order, and that order goes out to all the systems in the building, and they have to go find that inventory, and they have to pack it to a bin.
• 17:39
  So this, this robot's optimized for a plastic tote or bin.
• 17:43
  Um, when a human gets that, y- the, the orders have to be accurate, like one tube of toothpaste, one toothbrush, you know, one pack of crayons, but they don't have to be perfect'cause the human has a brain, and the human's gonna go fill in the gaps of whatever the w- the work order was.
• 17:58
  Or if somebody moved the box of, of inventory, they'll just figure it out, right?
• 18:03
  With Agility, it's not gonna re- feedback on an incorrect placed item.
• 18:07
  It's just gonna error fault, right?
• 18:09
  Yeah.
• 18:09
  So what has to happen- Yep, move on is those, that worksheet that Colin described that would typically go to the worker, in ZonePick they have a little h- uh, handheld or a watch or a headset that's telling them what to do.
• 18:20
  Now you have to translate that to what's called a mission structure.
• 18:23
  So that's a translation of software to get very precise and real time so the robot is given a specific mission, and then it can report back,"Yep, I completed the mission." And that's how you do an inventory reconciliation.
• 18:35
  It's how you verify the accuracy of the order.
• 18:37
  Yeah, to, to make that translation from the WMS or the ERP, that's what we're talking about.
• 18:42
  That's like, yeah, I agree with you.
• 18:44
  That's like a$5 million package if you go with a consulting house.
• 18:48
  Now, there's cheaper ways to do it.
• 18:49
  You and I would spend a fraction of that.
• 18:50
  Oh, yeah, absolutely.
• 18:52
  But if, if you are, uh, you know, this is gonna get to your nobody got fired for hiring McKinsey type of thing.
• 18:58
  Um- Yeah you know, people aren't gonna necessarily go the cheapest route when they, somebody is, you know, angles.
• 19:04
  Totally.
• 19:04
  That's a whole nother animal.
• 19:05
  'Cause you're, you just involved the CIO, right?
• 19:08
  So you went from a, a facility operations function to now you touched the inventory, so you have to go to the CIO, and then you also may have to talk to the CFO.
• 19:18
  So as that circle grows, that's how Deloitte, McKinsey, Accenture, all those companies find their way in,'cause they're already the approved vendor, right?
• 19:26
  It's the easy choice.
• 19:28
  And I, I think one of the very interesting pieces will be just on the actual warehouse operations.
• 19:34
  It, setting aside the actual robot piece of it- I, as much as everybody likes to believe that they have beautiful, perfect warehouses, stuff gets moved, you know, RMAs get put, you know, when, when returns go into the wrong bin, i- they are very rarely perfect.
• 19:53
  That's right.
• 19:53
  Part of the benefit of the human is they can see and swap things.
• 19:57
  Granted, it's also part of the reason why there are those issues, but y- that, that means you are going to be spending a ton of money getting your warehouse...
• 20:07
  You're, you're gonna be, uh, doing all of your physical counts again before, making sure everything's in the exact location.
• 20:12
  It, it's going to be a major undertaking just so that you have the underlying data structure corresponding with the real world before you can deploy these.
• 20:22
  Yeah, so- Because they're gonna be useless i- if it's a mess.
• 20:25
  Yeah, so, so just thinking about the economics and, and, uh, if you kind of take my numbers at face value, that it's a s- a six plus year payback, uh, the way the CFO would evaluate it, and then you, let's say that, um, because they have plenty of headroom in the margin, Agility's doing 50% off coupons for Amazon and Target to get that early customer adoption, that's that kind of buy the customer thing.
• 20:48
  Yeah.
• 20:49
  Um, all right.
• 20:51
  Well, you're still at a three-year payback, right?
• 20:54
  And a three-year payback in e-commerce s- still doesn't fly.
• 20:57
  It's, uh, they want stuff that's inside of one year.
• 21:00
  If it's a really compelling project, they'll consider maybe a year and a half to two years.
• 21:04
  Nobody's paying three-year paybacks, though, in e- in e-commerce, or three fails.
• 21:08
  So- I- So now- I just don't see it.
• 21:11
  With how quick things are evolving, too- Yeah, it's crazy what's gonna be available in three years?
• 21:15
  W- well, here's the next challenge.
• 21:17
  So the, the system, whether you rent it or you buy it, it's got a five-year shelf life, right?
• 21:22
  So they say.
• 21:23
  I mean, it's not field tested to know that, but they say it's got a five-year shelf life, useful life.
• 21:28
  Okay, so at the end of five years, I'm gonna re-incur those costs, right?
• 21:33
  I'm gonna have new costs, either to replace the robots or to continue the license fees on the software, whatever it is.
• 21:41
  Why not, and this is, this is gonna be the thing that you're gonna hear every VP of Ops say ac- around the country.
• 21:47
  Why not just buy an Autostore?
• 21:50
  Then if I buy an AutoStore and I delete...
• 21:52
  The reason you didn't do an AutoStore the first time was'cause you could get away with grocery s- grocery store style racks or shelving where you did your zone pick, and it was low cost,'cause all you had to do was buy the bins and the racks and put$18 an hour humans out there to do the picking.
• 22:09
  Okay.
• 22:09
  Well, here's the problem.
• 22:11
  Takes up a lot of space.
• 22:13
  The ceiling in that e-com facility's at least 30 foot clear, maybe 40 foot clear, right?
• 22:18
  So you're just wasting all the headroom.
• 22:21
  And, you know, some companies do- But you don't have to, do you?
• 22:24
  You don't have to if...
• 22:26
  So if you went with, like, an AutoStore, which is a dense storage, you know, bin system, or one of the...
• 22:31
  There's, you know, four or five, uh, competitors out there that are similar.
• 22:36
  I'm taking up maybe 20% of the footprint to get the same inventory, so now I've just freed up a ton of space and I've gone vertical and high density, and that can be easily integrated to your WMS.
• 22:48
  That already exists.
• 22:48
  That software exists already to take a WMS directly to an AutoStore or an Exotec or a mini-load system.
• 22:56
  The...
• 22:56
  You're, you're competing with traditional legacy automation, which is very well-suited for this, and there's many service providers.
• 23:04
  I wanna say AutoStore is up to...
• 23:06
  I use them'cause they're the number one brand.
• 23:08
  AutoStore is up to, like, eight integration partners in the United States.
• 23:12
  Wow.
• 23:12
  So that's eight different companies you can call to get a quote, you know, eight different companies that can service you on site, provide the spare parts, come out and do the software patch for the field deployment.
• 23:21
  You don't have the vendor lock.
• 23:24
  You're very, way less vendor locked if you go with one of those systems, and you get the footprint back.
• 23:30
  You could install more high bay racking.
• 23:32
  You could install a, a variety pack station where you can do promotional items, like, whatever it is you want it to do.
• 23:38
  So the only way I see Agility working is they're gonna have to discount the costs to probably no more than 100K a robot.
• 23:46
  That's probably the max threshold that, that this can bear, because we have to think about the ERP upgrade.
• 23:52
  We have to think about the software.
• 23:53
  We have to think about the workflow changes with rejects and returns and misplaced items.
• 23:58
  All that comes into play when we automate.
• 24:01
  So what I think they're gonna do is they're gonna sell this thing for 100K a robot to the big name customers.
• 24:07
  So you can say,"Look, Walmart is using it.
• 24:10
  Look, Target is using it," right?
• 24:12
  And they'll let it go for 100K.
• 24:14
  The business case works there.
• 24:16
  The, the next challenge is the engineering FDE, the forward deployed engineering labor.
• 24:23
  I mean, these things have to be installed.
• 24:25
  They they have to be integrated.
• 24:27
  The workforce around it has to be trained.
• 24:29
  You gotta get out to site, man.
• 24:31
  So I- You have to map the entire facility it's supposed to operate in.
• 24:34
  They...
• 24:34
  I mean, they must have a ton of employees.
• 24:35
  If you're burning 100 million a year, I can't fathom burning 100 million a year.
• 24:40
  It, it, it is, uh, uh, shocking.
• 24:44
  Um, uh, let's see.
• 24:46
  So- I'm scrolling here quickly.
• 24:50
  I'm not seeing Anything.
• 24:56
  Um Yeah, so I, I think what I read too was, just to, to now to go on the forward-looking basis, they had, uh, what they're calling$300 million in backlog, which again, that's probably those LOIs.
• 25:08
  I, I have no doubt that real companies signed real letters of intent that said something like,"Up to 200 robots.
• 25:15
  Up to 500 robots." Yeah.
• 25:17
  Right?
• 25:18
  Those are actually fairly easy for a, a well-formed startup to go after, and to get those, especially at the promise of kind of humanoid innovation, which is good for any CEO to say that they deployed humanoids.
• 25:31
  It sounds great to say,"Yeah, we're on the, we're on the cutting edge of doing this stuff." Yeah.
• 25:35
  "We're tech-forward." So, so the 300 million, you gotta take the discount on what, what is the actual price these things will move at?
• 25:43
  I think it's 100,000.
• 25:44
  I think that's the max ceiling.
• 25:46
  They'll move at 100,000 bucks.
• 25:48
  Here's the challenge.
• 25:50
  I'm burning 100 million a year.
• 25:52
  That would appear to be a structural part of the business, so every year going forward, we're gonna burn 100 million a year through R&D and marketing or whatever they're doing with the money.
• 26:03
  Um, now I'm adding to that a lot of payroll to increase my overhead,'cause I have to surge up my FDE staff to go- Oh, yeah deploy all these things, and then I'm only bringing in a contribution margin.
• 26:16
  I say only, it's a good margin, but 20,000 per robot, right?
• 26:20
  So I'm bringing back into the biz 20,000 a robot.
• 26:23
  The number of robots required to get break even, assuming no additional expense at building out your US manufacturing site, which you're gonna have to do, is$100 million divided by 20,000.
• 26:39
  That's the minimum break even number for the deployment of this.
• 26:44
  So are you beating me to it?
• 26:45
  Are you doing that math before me?
• 26:46
  No, I, I was looking at that 300 million backlog, you know, is 1,500 robots grossing 20,000 on them.
• 26:56
  That puts you at 30 million.
• 26:58
  No, yeah, so, so to make that you're gonna have to sell 100 million divided by 20,000, 5,000 robots.
• 27:05
  Right.
• 27:05
  So you're gonna have to- Right, I, I'm saying that- Per- 300 million backlog that they're saying they have wait for the follow, 5,000 robots per year, to break even.
• 27:14
  That's, that's more than triple what they have in backlog.
• 27:16
  And that's without the facility expense.
• 27:18
  That's gonna, that's gonna be another huge use of cash.
• 27:21
  And, uh, so that's 5,000 robots a year to get to break even at what, you know, my understanding of, uh, of facility economics.
• 27:30
  And, um, a- and in that time that it takes you to ramp to 5,000 and you've grown the org significantly with the, the e- extra deployment capabilities, then to your point, who comes along next to you that says,"We'll do it for 79,000.
• 27:49
  We'll do it for 49,000." You know, Unitree- And we can do X, Y, Z more.
• 27:55
  Yeah, I have no doubt that if Unitree built a replica of that, they would, they would want, like, 36K or something for it, you know?
• 28:02
  And so- I- that, it, it just doesn't work.
• 28:06
  Like, let's be honest.
• 28:06
  Uh, I...
• 28:06
  And, and that's kind of where I'm struggling, you know, generally with the, the, the, this robot as a service model.
• 28:13
  On top of that, if you think through, you know, if you're comparing, okay, I could go with your, you know, y- y- your improved facility, higher density, smaller footprint type of thing, there's also the benefit of, hey, we can capitalize over a longer duration.
• 28:32
  The, you know, we can capitalize this project and recover over a longer period of time with depreciation for that investment.
• 28:38
  Against this higher density, smaller footprint, more capabilities, versus every five years I'm riding, you know, 500K plus per robot that does- Yeah whatever.
• 28:51
  I, I'm skeptical.
• 28:54
  So I, I would say it's- You know, I, I, I would love to be wrong.
• 28:56
  I would love to be wrong.
• 28:57
  I'm just, I- I am very skeptical on the real opportunity for these to be effective.
• 29:05
  Yeah.
• 29:05
  I'm, I'm not skeptical, I'm just saying with that pricing, no way, right?
• 29:09
  No way.
• 29:09
  And I, I don't understand why they have so much overhead.
• 29:12
  Like, I, I didn't see$100 million worth of overhead- In that page you just had up that it, you know, uh, that's, that's kind of crazy to me.
• 29:20
  But so the, the next thing that we have to think about, you know, when we're considering like this cutting edge tech, we, you know, at my company, we underwrite this stuff all the time, right?
• 29:31
  I have all kinds of kind of Series B, Series C robot companies asking Chang Robotics to partner so that they can be pre-vetted for install in our facilities.
• 29:41
  Um, and then through our fund, the Chang Robotics Fund, you know, we get these same companies asking for investment and, and usually at the seed stage or, you know, kind of Series A stage of capital.
• 29:53
  So I've done a lot of underwriting on this type of technology.
• 29:57
  Y- you've seen a bit.
• 29:59
  And, and so here's one of the interesting things with RaaS,'cause if you look at it, if you, if you do a deep search with, uh, Groq, you'll probably find, you know, a couple dozen companies going after robots as a service.
• 30:12
  With software as a service, it's sticky.
• 30:15
  That's one of the things people have been surprised about in the age of AI, that yes, AI can rewrite the SaaS for you, but a lot of organizations are already plugged into their SaaS and the annual fee is just not high enough to justify changing.
• 30:28
  So the SaaS- Nobody wants a headache the SaaS subscription revenue has not really dropped as much as you would think with the emergence of, you know, Claude and all these other coding tools.
• 30:39
  So software is sticky, right?
• 30:43
  Nobody wants to change their software, it's pain in the butt.
• 30:46
  With, with robots because they're in your face on the factory floor or the fulfillment center floor, if it doesn't work, the workforce starts to revolt, the operations team starts to revolt.
• 30:56
  They just don't want it.
• 30:57
  They rise up.
• 30:58
  And it's literally- They rise up against the machines get that junk out of here.
• 31:02
  So when I've gone into other facilities where another robotic installer went ahead of us, I've seen robots stretch wrapped and put up in the, the shelves in those warehouse.
• 31:13
  The maintenance team just ripped them out.
• 31:14
  They're like,"Get these things out of here.
• 31:15
  We'll just do it ourselves." So RaaS is not sticky.
• 31:21
  The point I was making with underwriting, RaaS, I have seen RaaS contracts that got terminated in the third month.
• 31:29
  Wow.
• 31:30
  Like, like large orders where they'll sign up for a rental, they'll get them out there.
• 31:36
  It doesn't work as good as they said it would.
• 31:39
  In e-commerce you have this little thing called peak.
• 31:42
  Peak season is coming.
• 31:43
  You know, we got to prep up for our big s- actually we're about to be into peak here, uh, very shortly.
• 31:48
  Yeah, the holidays are coming up.
• 31:49
  So every, every fulfillment center in the nation is getting ready for peak right now, and that's the peak orders that come with Black Friday, Cyber Monday.
• 31:56
  And, uh, and so peak's on the way.
• 31:59
  You make the hard call, get this thing out of here.
• 32:02
  So RaaS subscriptions are not sticky at all.
• 32:05
  And then let's get to the deep finance side.
• 32:08
  I tried to bait Robert Stirling into weighing in on this and he, uh, he wouldn't take the bait.
• 32:13
  Uh, here's the challenge.
• 32:16
  If that RaaS subscription is over 12 months, which is what the, the startup wants, is a long-term subscription.
• 32:23
  If the RaaS subscription's over 12 months, you have to put that thing on, on the balance sheet as debt.
• 32:30
  It goes on your debt ledger.
• 32:31
  And so the CFO might be...
• 32:33
  Most CFOs are sensitive to debt because that debt to income ratio is, is important for share price So they're looking at it and saying,"You guys can't sign up- And they don't want to breach any finance covenants.
• 32:45
  Yeah.
• 32:45
  You can't sign up for debt.
• 32:47
  This-'cause it, it...
• 32:48
  then after 12 months it turns into a lease, and by and large, most leases are debt.
• 32:55
  And so all of a sudden, this thing that seemed like a great operational enhancement is giving the CFO headaches.
• 33:01
  You scale that out, you're scaling his debt, right?
• 33:04
  And that doesn't feel good.
• 33:05
  You're gonna get pushback on that.
• 33:07
  To your point, why not do a CapEx upfront and have the amortization over the useful life?
• 33:13
  Your competition at the same price point, like AutoStore, is the same price point.
• 33:20
  Um, that's a, you know, 12, 15-year useful life, you know, versus five.
• 33:26
  Mm-hmm.
• 33:26
  So at the CFO's desk, it's not gonna fly, man.
• 33:30
  I'll just tell you right now.
• 33:31
  So my advice to Agility, I, I don- I doubt they want it based on the momentum they have, but my advice to Agility is focus on minimizing your cogs so that you can get that robot out the door for the absolute lowest cost possible per unit.
• 33:49
  Next is start to develop a base of integrators or channel partners so that you don't have to absorb all that FDE engineering, that forward deployed engineering.
• 34:02
  You can rely on other companies that add you to their portfolio.
• 34:06
  Maybe they already install AutoStore.
• 34:08
  Maybe they already install sortation systems.
• 34:10
  Maybe they already install robot arms.
• 34:13
  And so you can use them, because there's probably gonna be a project in their backyard.
• 34:17
  It's gonna be cheaper with better customer service if they just go do it.
• 34:21
  What that means is you gotta get that cogs down so that you can have a fair profit, not a 66 or 70% contribution margin, but a fair profit.
• 34:33
  And then your channel partners can start to handle those really weird integration points that you're gonna have.
• 34:39
  These guys are g- if they, if they scale to 5,000 units a year, they're gonna touch probably 20 enterprise software systems.
• 34:48
  Oracle, SAP, Blue Yonder, and every other flavor of ice cream out there.
• 34:55
  What a headache, man.
• 34:57
  Why?
• 34:57
  It's, uh- That's my question.
• 34:58
  Why?
• 35:00
  I, I...
• 35:00
  you know what?
• 35:01
  At, at the end of the day, I, I'm not rooting for anybody to lose.
• 35:04
  I would love to be completely wrong and to see it succeed.
• 35:10
  I, I am simply skeptical.
• 35:13
  To your point, the economics are hard to see how the business justifies those- Might be a little harsh, but I think anyone who kind of goes all in with something like this at this stage is a bit performative in trying to...
• 35:28
  They, they see more value in the headlines and the press releases than they do the actual product itself.
• 35:35
  They're spending 46 million a year on, you know, their admin overhead basically.
• 35:43
  You know, that's, uh, y- you, you shit- Wow, you went, you went deep on the numbers.
• 35:47
  Good job, Colin.
• 35:48
  Y- y- you shit.
• 35:49
  What, y- I, I was curious, you know, is, you know, f- sorry, 46 on SG&A, 92 on R&D.
• 35:58
  There better be- Their cogs- an airplane to come with that.
• 36:00
  Oof.
• 36:02
  W- yeah, I mean, their cogs were, you know, 4.5 million- Oh, man on 1.8 in sales.
• 36:09
  Well, that, what that shows you is the discounting, right?
• 36:13
  Yeah.
• 36:13
  So if you got 4.8 million, uh, cost of goods sold, and you got 1.8 million revenue, that, we just call that buying customers.
• 36:21
  You're, you're getting- Oh, yeah pilots out the door for a hugely subsidized...
• 36:25
  Using investor capital to subsidize the cost of a deployed robot.
• 36:30
  But I go back to it, are they manufacturing and assembling here?
• 36:34
  Do we know that?
• 36:35
  I, I've not been able to see, uh, been able to find anything definitive, um- It, I mean, it didn't matter'cause they, they only shipped 18 robots according to that- Yeah report they did.
• 36:46
  So it didn't matter.
• 36:46
  You could do that with the university guys, uh, that they came out of, right?
• 36:51
  They can self-assemble.
• 36:52
  It's like the drone companies are like that.
• 36:53
  They can build up to 20 themselves, and then after that they have a problem.
• 36:57
  Uh, yeah, to get to 5,000 you're gonna need a full on assembly plant with, wait for it, your own inventory systems.
• 37:07
  So it's back to that you'll know humanoids work when you see the humanoid building the humanoid.
• 37:12
  You'll know Agility- Yeah works when you see an Agility assembly plant, and Agility's doing many of the tasks inside the plant to build the robots.
• 37:21
  I don't know.
• 37:22
  We'll see.
• 37:22
  I, I'm, I'm rooting for everyone to, to win.
• 37:26
  You know, I, I wanna see the, the needle move, if you will.
• 37:30
  I'm just...
• 37:33
  I'm not a fan of SPACs, uh, so seeing, you know, 2.5 billion- Oh, oh go down that.
• 37:38
  I, I wouldn't judge them for that.
• 37:39
  I mean- No, I'm simply saying s- You know, that, that part I've got, I've got grace on.
• 37:44
  Here's, here's the way I read it, is they're doing the$2.5 billion, uh, is the target share price, 2.5 billion.
• 37:53
  Um, and they're, they're doing the SPAC.
• 37:56
  Most SPACs don't have any money in them, right?
• 37:59
  Yeah.
• 37:59
  So if the SPAC, if the SPAC tells you,"We've got 100 million or 500 million," they, they redeem, they have a redemption right at the closing of the SPAC, at the closing of the, the merger or the acquisition.
• 38:12
  So mo- more than 90% of the capital will go back to the original investors of the SPAC.
• 38:20
  Very little of it will sit in the company.
• 38:22
  And the reason is, um, the SPAC investors get a warrant for every share of the SPAC they had.
• 38:29
  Okay.
• 38:30
  So it's like I can take my cash out with no risk, and what I've earned is one share or one warrant per share of the SPAC that I have that I get for free because I funded the SPAC up front.
• 38:43
  So I...
• 38:44
  So there's gonna be, my guess would be, mm, 10 to 30 million bucks left in the SPAC, which is only three months worth of runway for Agility.
• 38:54
  So they're expecting 400 million.
• 38:57
  Ooh, ooh.
• 38:58
  So let's see, it says- Not gonna happen they're, they're expecting four- Not gonna.
• 39:01
  Let's see.
• 39:02
  Yeah.
• 39:02
  Expected to generate more than 620 million in gross proceeds, 420 million from the, uh, was it Churchill Capital Corp, th- th- they're, who they're merging with.
• 39:14
  Yeah, that's what I'm saying.
• 39:16
  Yep.
• 39:16
  Church- Churchill has 450 million in the SPAC today.
• 39:20
  Yep.
• 39:20
  The shareholders of Churchill are not gonna leave their cash in Churchill because they already have a one-for-one warrant in the acquisition.
• 39:28
  So they're just gonna pull the cash out, redeploy it elsewhere, roll over the warrants into Agility, which is the smart play.
• 39:36
  Yep.
• 39:36
  And so there's only gonna be 10 to$30 million left in the SPAC, okay?
• 39:44
  Then when the merger completes and it become, the whole enterprise becomes Agility, at that point in time, I also read there was like a$200 million pipe, uh, which is the private invest- Yep, from Foxconn with the private investment.
• 39:57
  Oh, Foxconn's in here.
• 39:58
  Are they the manufacturer?
• 40:00
  That's what I'm wondering.
• 40:01
  I, I can't seem to- It's gotta be I was doing a little search and I couldn't seem to find it.
• 40:04
  Um, but the- Hey, hey, here's the silver lining$200 million pipe led by Foxconn.
• 40:08
  Here's the F- here's the silver lining.
• 40:10
  Isn't Foxconn in Lordstown, Ohio and in Wisconsin?
• 40:15
  Uh, they are.
• 40:15
  They actually, not too long ago, I guess earlier this year, late last year, they acquired, um, one of the facilities in Lordstown, the one that went through a couple different hoops.
• 40:24
  Yeah.
• 40:25
  So m- hey, maybe they build this thing in Lordstown.
• 40:27
  That's the, your silver lining right there.
• 40:29
  Absolutely.
• 40:29
  In that case, we gotta change our tune and cheer for it,'cause I'm all, I'm all in on Lordstown.
• 40:33
  I, I, I am cheering for it.
• 40:34
  I'm just, I, I, I'm skeptical.
• 40:36
  So- You know, I want, I want to be wrong.
• 40:39
  So let me...
• 40:39
  I'm j- I'm just being realistic here of what the actual usable capital is.
• 40:44
  So they're gonna, they're gonna get the$200 million from the PIPE, right?
• 40:49
  That's the investment into the, that goes into this...
• 40:52
  It's kinda like the pre-IPO, the pre-SPAC investment, the 200 million cash.
• 40:57
  They're gonna keep, let's be generous and say 30 million from the SPAC, will stay in there.
• 41:03
  So they're gonna start off post-SPAC with, uh, 230 million fresh cash.
• 41:10
  The challenge is, Colin, that's a two-year runway.
• 41:14
  18 months.
• 41:15
  This, yeah, it's gonna go fast.
• 41:19
  And so as they advertise a$300 million backlog, you better be able to jam all that revenue basically into year one.
• 41:29
  But it's a five-year...
• 41:30
  The 300 million is a five-year forecast,'cause look at the pricing sheet.
• 41:33
  It's all based off a five-year plan.
• 41:36
  So, you know, uh, w- let's be generous.
• 41:40
  They got two years of runway when you include the new revenues and you include the, uh, the PIPE and, and a little residual in the s- in the, uh, SPAC.
• 41:51
  Uh, you know, could you make it?
• 41:53
  Sure.
• 41:54
  I mean, I, I think I could, could probably- I could probably run that company with$5 million worth of overhead.
• 42:03
  I can't get to the$100 million.
• 42:04
  The over- the overhead absolutely blows my mind.
• 42:06
  Like, that number is just absolutely insane.
• 42:08
  Well, there was another famous company that did that too.
• 42:12
  Uh, if you remember Attibotics out of Canada, they did that.
• 42:16
  They had really kind of anemic sales.
• 42:20
  Uh, fantastic product.
• 42:21
  The, again, the engineering was incredible.
• 42:23
  The, uh, the innovation was incredible.
• 42:26
  But their burn rate relative to their revenue was, like, 10X, 10X overhead to the revenue, and it just wasn't sustainable.
• 42:35
  And in their story, they missed one funding round.
• 42:39
  Oof.
• 42:40
  So they were, they were out past series D or something like that, and then the, the final funding round that needed to happen to kind of get them to where Agility is now, to this kind of SPAC level, uh, blew up.
• 42:51
  So here, here's the reason why, uh, if you didn't know, this...
• 42:55
  You got me so deep in the rabbit hole.
• 42:58
  I wonder if anybody wants to guess, why do we have SPACs?
• 43:02
  Like, what is the real reason that a company would SPAC?
• 43:05
  And the short answer- I mean, that- Go ahead.
• 43:07
  It, it's a lot easier to, to be acquired or to merge with a company than it is to go through the audit of all the financials, the compliance- Boom SOX, all that good stuff.
• 43:16
  Boom.
• 43:16
  You nailed it.
• 43:18
  It's because you can't get audited.
• 43:21
  That's why.
• 43:21
  So what's- You don't have clean enough books.
• 43:23
  So, well, it's- You don't have controls, you don't have, you, you don't have all the things that you typically need for public company.
• 43:30
  'Cause you, you give up, you give up about 20% dilution when you do a SPAC merger.
• 43:35
  So to the existing shareholders and the founders and all that, that, that share the cap table, you're gonna collectively give up about 20% for the SPAC merger, right?
• 43:45
  And that's, that's before the investment.
• 43:47
  The investment in this case is another 10% or so.
• 43:50
  So they're giving up probably in the neighborhood of 30% dilution for the SPAC merger.
• 43:57
  But the reason you SPAC and the reason you'll pay 20% is because you don't have three years of trailing financials, and you can't get a two-year public company audit on the bids.
• 44:07
  So at that point in time, a- and that, the, think of a high-growth startup that at 18 months is at breakout velocity.
• 44:15
  We see those sometimes You need access to the capital markets Then a SPAC makes all the world of sense,'cause you're not gonna sit there and wait another 18 months.
• 44:22
  You're already at breakout velocity.
• 44:24
  Why wait?
• 44:25
  Go ahead and SPAC and, and move on with life.
• 44:27
  So that's why I say SPACs are not always bad, but it is a...
• 44:30
  When you see a SPAC, it is a telltale sign that the company can't stand on its own two feet, especially if they've been around 10 years and they can't pass the audit.
• 44:42
  That's concerning.
• 44:45
  So I- here's, let's wrap it up on a high note.
• 44:48
  Love the form factor.
• 44:50
  From the videos I've seen, I love the speed.
• 44:52
  The, the opportunity charging I think is just insanely cool.
• 44:55
  Uh, insanely cool.
• 44:56
  I'm gonna have to check out the charging.
• 44:58
  I'm, I'm certainly excited to see that.
• 45:00
  That's what I look for.
• 45:00
  And despite my pessimism, I- That's what I see the eyes, too, eyes, too, are cute.
• 45:04
  All right.
• 45:04
  Despite my pessimism, I am in fact hopeful that they are successful.
• 45:08
  And now I am rooting for them based on the theory that they're going to partner with Foxconn and manufacture these robots in Lordstown.
• 45:17
  I hope that's true.
• 45:19
  Let's hope.
• 45:19
  I'm gonna have to dig into that more.
• 45:21
  I'm curious now.
• 45:22
  You've got my, uh, my, my gears turning in my head, so I've got more questions now.
• 45:26
  I won't write it off.
• 45:28
  Yeah.
• 45:28
  But yeah, awesome episode.
• 45:30
  Thank you as always.
• 45:31
  To all of you listening, watching, tuning in, thank you for another episode of the Reforge podcast.
• 45:37
  As always, whether we're talking the latest reindustrialization news, robots on the factory floor, policy that moves the needle, or catching up with people rebuilding America's industry, our goals are always to show you that reindustrializing America is possible.
• 45:51
  It's happening right now, and it's pivotal.
• 45:53
  Make sure you subscribe and follow the Reforge podcast on all of our socials, and be sure to check out thereforgepodcast.com for more, including our seven-step guide to reindustrialization.
• 46:03
  Until next time, keep building.
• 46:10
  Thanks for tuning in to the Reforged podcast.
• 46:12
  Make sure to like, subscribe and follow us on all of our social media@thereforgedpod.
• 46:18
  We will see you next time