Requiem for an Innovation: What We Can Learn from the Sad Story of Laundroid, the Laundry-Folding Robot

Our society tends to narrate innovation in triumphalist and only triumphalist ways. As we, particularly in the West but increasingly as a global society, have collectively embraced the notion of innovation as something always already good, always already right, always already the one prime thing to aim for and desire, this should come as no surprise. Tales of innovation tend to be tales of heroes and progress, better lives and happy customers, with a promise of greater things yet to come. Why not sing the praises of a thing that seems like such a horn of plenty? Yet this is a very narrow and blinkered way to talk about innovation. Every child knows that for every innovation that conquers the world, there is a plethora that falter and fail, for various reasons. For this reason alone we should pay at least as much attention to those cases where an innovation stumbles and falls as we do to these who succeed, but we do not. This is a shame, for this is part of why we fail to see many of the cracks in the valorized edifice of innovation, the many issues hidden from us as we stare intently at only the visible, successful top of the innovation iceberg. In the following I will present not a full attempt to rectify this situation, but merely a sketch of a singular instance of innovation failure, to show what can be learnt from this. It will be a requiem for a robot, but this sad funeral dirge isn’t just about remembering what has passed — it is a reminder that just because something is called an innovation, or might even be seen as an innovation, this doesn’t mean that what we missed was a great, good thing. Sometimes things fail for a reason.

In the years leading up to 2020, robotics has been one of the hottest and most hyped technologies. The heat and hype of course continues, and not without reason, but we can also see that some of the robotics companies that were meant to rock our world have now shuffled of their mortal coil, or to be a little less grandiose, quietly folded. Among them Laundroid, a company that well deserves to be the poster child for robotics hype. Perhaps more importantly, it is a very educational tale about innovation in our age, where novelty and shiny demos often beat out purpose, impact, or even sense.

Robots can be used for many, many things, and they increasingly are. We’ve gotten quite used to robots that mow grass and vacuum flats, and more and more often we run into robots in other areas of everyday life — scuttling about in a warehouse, making a pizza, or delivering food from Tescos in Milton Keynes. Heck, the next time you go get a drink, your bartender could be a robot. No wonder, then, that a lot of people have grandiose visions about all the things that could be delegated to our handy, helpful robot companions. We’re still a long way away from general-purpose androids, i.e. robot butlers who can take care of all of our little everyday chores, but this hasn’t stopped people from trying to bring this dream about piecemeal. One small part of this is to design a robot that could take care of folding laundry, a task that few people cherish but which must still be done with annoying regularity. Laundroid to the rescue!

The robot known as Laundroid was a much hyped product, and for a few years a mainstay at CES (the massive trade-show of the Consumer Technology Association). An early prototype was demonstrated in 2015, to an intrigued audience. The sales pitch for it stated that it would eventually develop into a full-fledged clothes-processing robot, which would be able to take a load of washing, wash it, and then dry, iron, sort and fold the clothes, depositing them in a closet built into the device. The early version would however only be able to do the last part — fold clothes and put them into a closet. The technical demands of even this less than thrilling function shouldn’t be underestimated, however. It requires computer vision, a smart system to understand what the robot is folding, and robotic arms to fold garments without doing damage to them. As a result, the Laundroid was neither small nor cheap. In fact, it was remarkable for both its bulk and its price tag. Size-wise it was big as a small fridge, if a bit deeper, and thus not something that would be easy to fit into a small bachelor-pad — or many family homes. However, the price of the robot did make sure that this wasn’t really a problem. This, as the price was set at 16,500 USD, pricing out the kind of people who would have an issue with not having enough space for it.

Who would pay 16,500 dollars for a machine the size of a fridge that folded laundry? As it turns out, not a lot of people. The kind of people who could afford it could also afford either a maid or to have everything dry cleaned, and the path towards the additional functionality proved to be problematic. Whilst I am sure a lot of people would be thrilled to be able to dump a load of laundry into a robot, and get it back clean, ironed, and folded the following morning, this merely compounds the already tricky technical issue of getting a robot to fold clothes properly. How do you do this in a machine that needs to deal with water (for washing), high temperatures (from drying and ironing), vibrations (from washing and drying), all whilst keeping track of robot arms and a bunch of optics?

In the end, Laundroid never got that far. On April 23rd, 2019, the company developing it went bankrupt, having never shipped a single unit. It turned out that the complexity of getting AI, computer vision, and robotic arms to work reliably in concert made the problem too difficult to reliably solve. A reporter from Verge reported that the demo machine he tested in 2018 couldn’t handle folding his t-shirt, simply because its black color confused the image-recognition system (because, you know, black is such a rare color for t-shirts…). In other words, the Laundroid failed, for reasons both technological and market-related. Now, here’s the truly important bit. The company behind Laundroid, Seven Dreamers (registered in Tokyo) had according to reports raised a hefty 89 million USD. Not only that, they ended up in debt to the tune of an additional 20 million USD. Let us repeat that for emphasis:

More than 100 million USD was given to and spent by a company to build a laundry-folding robot.

Regardless if you feel folding your laundry is a chore or not, realizing that amount of money went into creating a solution for something that simply isn’t that big of a problem for most people should at least shake your belief in a fully rational innovation system. In a world jam-packed with actual, wicked problems, why would investors line up to fund a very expensive way to deal with what might best be described as an annoyance? The list of reasons (sadly) reads like a checklist for modern innovation: The Laundroid was a high-tech solution, it had an AI (artificial intelligence) component, it was directed to affluent urbanites, and it was just perfect for fun demos and PR fluff pieces. The fact that it was oversolving a non-problem really didn’t matter.

So now Laundroid is no more. The money put into it of course didn’t disappear, as they went to staff, PR companies, sub-suppliers, and so on. At the same time, a lot of assumedly smart people spent time on something that never did come to fruition, and whilst the money was kept in circulation in the economy, they were also kept from other, potentially more worthy innovation projects. To me, the Laundroid is an interesting story of shallow innovation, innovation that whilst it might be novel and flashy was never bound to have all that much impact. Sure, the apologetes of innovation will say that developments here might at some point have led to other, more worthy innovation, such as robot surgeons or other fine things, but such apologetics ring fairly hollow to me — as we simply cannot know whether the future path of the Laundroid would have gone in that direction. Further, it would take quite a lot of mental acrobatics to state that right here, right now, the thing we should invest time and money on — in the name of innovation — would be developments in laundry-folding. That dog simply won’t hunt.

No matter how you read it, the Laundroid wasn’t a triumph. What its failure teaches us depends on the perspective we take on it. Was it a moon- or loonshot that simply was before its time, or was it a wasteful boondoggle that shouldn’t have been started in the first place? There is no one, clear answer to this, yet this shouldn’t stop us from discussing the matter. In fact, we should talk about cases such as the Laundroid far more, to understand the amount of money invested into projects with unclear value and impact, and to understand what happens when the grandiose stories of innovation and technology do not quite pan out. So, a requiem for a robot. Laundroid, we hardly knew ye. Yet you too were part of the grand tale of innovation, and your story deserves attention. As a warning example, if nothing else.