Thanks for the links. Indeed, hilarious to watch it side-by-side for some extra entertainment! With the volume of pizza they're making in that Zume factory, it doesn't seem like there's even any extra benefits of using robots — and these robots seem to be placed there more as an obstacle to overcome than actual help; also, they didn't even use it for cheese dispersement, which was still done by hand? LOL, robots-made pizza!
The automation in that video seems less than helpful, all the labor-intensive steps are being done by the humans, all the super easy stuff done by robot.
The six-axis robot being used to move a pizza 2 feet was pretty amazing -- I have to imagine they had grander plans but had no idea how to utilize it yet.
The mobile ovens are kind of absurd. I mean, warming it doesn't seem like a bad idea. Maybe even baking it on the way is fine (so long as all the toppings didn't slide off), but what do they need 56 bays for? If they're planning on loading up 56 pizzas and baking them on the way, latency on orders will be horrible. Huge orders are like, 10 pizzas. Seems like if they just baked them and stuffed them right into a warming compartment that'd be more than sufficient.
I noticed the same thing! They just needed to move the pizza across a 90 degree angle so they used a robot arm? Why not just rearrange the conveyor belt a bit?
You want to be able to say you use $COOL_TECHNOLOGY, so you use it to move a pizza two feet and do all the actual work the traditional way. It's like AI or blockchain, but for pizza.
>The automation in that video seems less than helpful, all the labor-intensive steps are being done by the humans
The saucing of the dough in particular was absolutely slower than a reasonably experienced human worker. For a human worker in a typical pizzeria, the sauce takes two seconds.
I know that with repetition comes this kind of effeciency, but that's amazing to me, because when I make pizza it still takes me, like 45 seconds to really get it done...
You can see how quick they move in a typical pizzeria right at the beginning of this video. They are slinging sauce so fast, I wouldn't doubt it is about 2 seconds.
The robot is incredibly over-sized and under-utilized for this task and costs what you would pay for 5 employees in a year - not the best application. If the robot isn't tossing the pizza dough, then who cares haha!
Something I love about the frozen pizza setup is just how much of it is about being sloppy and recovering gracefully. Cut rectangular dough band into circles and circulate the off cuts back into the start. Rain cheese everywhere and let it fall into a recovery tray. These are techniques that also work well in software — it’s often easier to design a system that builds a bad result and fixes it after than it is to build a good result from first principles.
The brilliance of the cheese process struck me as well – we see a mess and then, the reveal of the amusingly simple cleanup. (This is an effectively edited promotional film, too. The story of the equipment is beautifully told.)
Your insight re: the parallel to software design is a valuable one, thank you.
Zume eliminated the two easiest steps with costly robots using needlessly complex technology and that was the basis for their business? What insanity would prompt someone to invest in that?
I had family working in a plant for one of the big name pizza places, and you can see the whole process for on-demand pizza at many brick oven pizza joints. I don't know how you could go to either of those places, look at zume, then conclude that zume had any advantage or was solving a problem from either side (on demand or automation).
I would think that the biggest problem in this type of old commoditized industry (food) the biggest problem is how to sell. The cost of making lots of pizzas, with robots or humans, should be less in total than the cost of marketing, delivery, etc. Maybe that's the answer to this question. These robots are there to make these pizzas appealing to investors (hi-tech something something), and then prospective customers (a hi-tech pizza that doesn't taste like shit)...
Pizza in particular is fiercly competitive. There are tiny Domino's Pizza places in every city and all over the place. Their overhead is low and you can carry out a large pizza for $8. And it is not a bad pizza. It's not an artisinal pizza but it's not bad.
The most painful part is the founder lingo all this technical superlative 6-axis factory line robot [0] and laser triggered faucets just to make your sad blind ambition sound valueable. Also patenting ovens in a van. Icing on the cake. Or should I say robotic top filling on the automated dough.
To honest though, they look like fresh pizzas. And if by some chance, the system yields stable long term value (happy employees and eaters) then.. why not.
That said anybody gets depressed by the dehumanizing aspect of this era ? drone delivering robotically made pizzas.. it looked funny in the jetsons but in reality I'd like to keep the good'ol restaurants with noisy people in it.
[0] "that we have yet to master" because it's so useless and we're so unqualified to understand it
Yes, that's nuts. That super expensive 6 axis robot is for what? If they just moved the oven, the pizza could transition to a wire conveyor right into the oven.
Not only that. There's already a pizza on demand system.[1] "Let's Pizza" has been around since 2011. From Italy.
Modestly successful. Softbank didn't do enough due diligence to find that?
Ah, that's disruptive I guess? You know, because old pizza is the non-digital incumbent and such? No idea, the more I think about it the more I prefer to not sell my business as a start-up. Because of stuff like that. And that dog walking company SoftBank invested in.
Perhaps that is the general difference between industrialization and AI based robots
Industrialization has pretty much solved mass production.
AI can automate intelligent decision making, but with industrialized production system the design has removed all need for intelligence.
An intelligent system will always be more expensive than a non intelligent system.
Robots could only become useful, once there are decisions to make. Like when you want to order truly customized pizzas, e.g., arrange the cheese such that it spells your name. Or give it a keyword, and then the robot should arrange all ingredients, such that the pizza looks like a painting about that keyword
Interesting trivia: Zume owns the patent for "cooking on the way". These days you can really patent everything (in USA only, I hope)... What about mobile hot dog stand? Isn't that prior art?
Pizza Hut had a 1984 patent for a mobile delivery/kitchen truck that would cook the pizza en route to the customer.[1] Zume's patent even cites it[1]. Just skimming over both filings, I'm kind of surprised it was issued in the first place.
I'm sure there's enough desperate money like Son's in the VC/PE world that a patent for anything and a decent story can get you a meeting and a term sheet.
Love the production line. I was especially impressed by the cheese application step.
I was like “oh these pizzas aren’t in a straight line, I wonder if a human needs to line them up now, or if there’s a visual sensor that detects edges”.
I don't think VCs care about that. In fact, they probably prefer that it has been done before. It's not like Amazon was the first company to sell things online. Nor was under the first to have an app to hail a ride.
> Sure, Rademaker doesn’t put their enormous frozen pizza technology in a van.
I bet if you gave them enough money to bother trying, they could squeeze a pizza machine into a van. Making a pizza in a van seems possible, but pointless.
How the hell can people controlling so much money be so ridiculously gullible? WTF did I just watch? A 6 axis factory robot who's sole purpose is to move a pizza 2 feet from a conveyor belt into an oven? Not only that but they're still relying on humans to actually place toppings? And they got $375MM for this bullshit?
Serious question: would this investment have been awarded if the cofounder weren't a minority female?
"Bruno is a very powerful robot. He's a 6-axis robot. You've seen this kind of robot mostly in manufacturing settings. We're using this robot to make pizza."
Especially since an oven is probably the simplest tech in the entire assembly line, and a relatively short distance (if not the shortest), so it seems kinda dumb to change the line's direction there, both in expense and complexity.
> A 6 axis factory robot who's sole purpose is to move a pizza 2 feet from a conveyor belt into an oven?
I have to believe they bought it with the idea of doing much much more but hadn't figured out how.
Overall, it seems they knew they wanted to "automate pizza" but really had no idea how. Another comment shows a real pizza assembly line that looks both simple and efficient.
Lets talk a minute about their "mobile ovens". WTH do they need 56 ovens for? It seems like assembly still has to happen at a "hub", but how many pizzas do you really need to bake on the way? Maybe a max of 10 for huge orders?
> Serious question: would this investment have been awarded if the cofounder weren't a minority female?
> I have to believe they bought it with the idea of doing much much more but hadn't figured out how.
The video gave me the impression of a kitchen designed by a chef or line cook, with a manufacturing robot tossed in there. If the goal was automated centralized production, any competent manufacturing engineer would have designed a radically simpler setup. Of course, finding machinery that didn't have a production capacity too far beyond Zume's needs would probably have been a challenge. Hundreds of pizzas an hour--let alone thousands--is well beyond anything Zume could have handled, because their biggest bottleneck would have always been delivery.
> Lets talk a minute about their "mobile ovens". WTH do they need 56 ovens for? It seems like assembly still has to happen at a "hub", but how many pizzas do you really need to bake on the way? Maybe a max of 10 for huge orders?
The entire concept is a bit wacky. You don't need to bake the pizza entirely on the way; just scale up how New York pizzerias handle pizza by the slice.[0] A plain pizza gets half-baked and when a customer orders a slice, toppings are added and it's put in the oven for a couple minutes until fully cooked. Prepping whole pizzas at a central kitchen and then putting them in a delivery van that finishes cooking the pizzas is something I could maybe see being viable with minimal investment. A very hesitant maybe.
But there are no robots and no wizards (er, machine learning some of Zume's early coverage discussed) for that kind of business. And that probably means no big investments as there are some articles where I can't help but substitute "magic" or "wizards" for certain words and think "yeah, that's probably closer to what some investors or journalists heard."
I saw an analysis of Shipley a couple of days ago. Apparently Shipley refused to hire people from the shipping industry. Because these people would refuse "innovation". Compare that to Amazon Logistics who hired a ton of people from other shipping companies. Maybe, just maybe, industry and domain knowledge are helpful. Even for disruptive start-ups.
Plus, many of those grizzled veterans would love to run with a green field project. They've been putting up with obvious inneffiencies for years because most of the time, it's not the experts that resist change, it's the inertia of a mega Corp that is doing so much business right now, that change is extremely difficult and disruptive.
Yeah, so true. Consider me one of them. Occassions like this, Shiply, often strike me as young, aspiring scientist using other peoples money to figure out the a wheel should be round.
Being a little bit more active in my local start-up scene now also means following stat-ups in general. And a lot of these companies, including one I used to work for for a short period, seem to ignore lessons other learned, the hard way. If these lessons are concerning your USP, ignore them, because it is where you inovate. regardless, you should have understood these lessons before ignoring them.
Ignoring these lessons in any other part of your company is just stupid. Like not managing your cash flow.
Very different. Frozen pizza factories are impressively efficient (with decades old tech).
Compare random frozen pizza machine https://www.youtube.com/watch?v=7anib2L7uUk with the hilarity that is Zume https://www.youtube.com/watch?v=uFSdxwRVh8A