Why do so many home robots still suck?

"The Jetsons" debuted September 23, 1962. The first episode, titled “Rosey the Robot,” was an origin story of sorts for the titular character, describing how an overworked Jane hired the housekeeper. Sixty-two years after her debut, Rosey remains an important pop cultural touchstone for the unfulfilled promise of home robots.

The home of 2024 isn’t necessarily wholly devoid of robots. According to its own figures, iRobot has sold more than 50 million Roombas. That, meanwhile, is a fraction of the overall number of robot vacuums that have been sold around the globe. Robot lawnmowers and pool cleaners have gained traction as well, though those figures pale in comparison to their vacuuming counterparts.

Home robots’ unfulfilled potential isn't because of a lack of consumer demand or lack of effort from manufacturers. It’s more complicated and nuanced than that, though ultimately it’s a matter of pricing, functionality and efficacy. Outside of the aforementioned use cases, today’s home robots don’t do enough or do what they do well enough, and building a robot that can tick both of those boxes would prove prohibitively expensive for those of us who can’t afford our own islands.

Vacuums make for good home robots

During his long tenure as iRobot CEO, co-founder Colin Angle was fond of saying that he didn’t become a successful roboticist until he became a vacuum salesman. It’s a fun quip that gets to something much deeper about the industry. Before the Roomba came along, the company had experimented with everything from baby dolls to military equipment.

iRobot found success when it focused on a simple task: cleaning floors. The earliest models were primitive by today’s standards, but they got the job done well enough to justify their price point. In addition to marking 62 years since Rosey's TV debut, next month is also the Roomba's 22nd anniversary. The robot vacuum is old enough to legally buy a case of Sam Adams.

In the nearly quarter century since the Roomba launched, much of iRobot’s R&D has gone into making the system smarter, adding sensing, mapping, and AI and integrating with smart assistants. The company has invested into other robotics categories as well, including gutter clearing, pool cleaning and a lawnmower that may never see the light of day, but all have failed to recapture the Roomba’s magic.

Work/home balance

Years ago, I appeared on a panel to discuss robots. The moment the conversation ended and the Q&A began, a woman’s hand shot up, front and center. She was eager to tell me about her billion-dollar idea: a drone that vacuums, dusts surfaces and does the laundry. I told her it was a great idea and I would happily buy one from her when she got it up and running.

Everything in robotics is easier said than done. It isn’t that no one before her came up with the concept for a furniture-dusting drone; it’s that no one before her figured out how to build a reliable and robust version at scale with a price tag that isn’t higher than my 30-year mortgage.

I thought back to that moment when Tesla announced its robot by way of a dancer in a spandex onesie. Elon Musk described a humanoid that would toil away all day at the factory and then pick up your groceries on the way home, before preparing your dinner. Musk has been in the public eye long enough to know precisely how much stock one should put in his timelines.

Before the fully functioning Roseys of the world arrive, simpler machines are going to have to pave the way. Robots have had a place in manufacturing for decades, but they’ve been built to do one job well over and over again. The more complex the machine, the more expensive it gets and the more potential points of failure emerge. Think about how many ways your Roomba has failed and multiply that by the complexity of a humanoid.

Most experts agree that early home robots will be designed for a handful of simple tasks: social robots and those providing caregiving and doing housework. For the foreseeable future, each will be designed with one or two functions in mind.

Laying the groundwork for future robots

There’s a sense of frustration that home systems are nowhere near where expected by this point in time. What regularly gets lost in that conversation, however, is the amount of groundwork that has already been laid. Whoever builds the next great home robot won’t have done so in a vacuum.

Their success will be built not only on top of ongoing research, but also on the home robots that came before. Navigating in an environment as unstructured and dynamic as the home likely felt impossible for many before the first Roomba arrived. Again, it was a simple machine by today’s standards, but it laid the foundation for what comes next.

One can see reflections of this in the current crop of home robots. Take Hello Robot. As design goes, it’s an extremely simplistic machine. It’s an arm attached to a pole attached to a Roomba-like base. Its simplicity is, in part, because it’s more development platform than product. But systems like this or, say, Matic’s robot vacuum, are continuing the hard work of building foundations, be they mapping, manipulation or navigation.

Antisocial social robots

Social home robots had a rough year in 2019. Anki, Kuri and Jino all fell in quick succession, each for a combination of price, limited functionality and reliability. More recently, Amazon’s Astro has been dead in the water, effectively kneecapped after Amazon's belt tightening wiped out a significant portion of the company’s consumer hardware headcount. Of course, had the $1,600 robot been a wild success, the retail giant almost certainly wouldn’t have let it die a quiet death.

Amazon’s struggles are a friendly reminder that being one of the world’s largest companies doesn’t guarantee success in such a treacherous category.

And following the Vision Pro’s lukewarm reception, one has to wonder whether Apple might be walking a similar road with its reported home robot play.

A more recent report suggested that the first project out of the group could look less like Amazon’s Astro and more like its Echo Show 10. The potential product has been described as something along the lines of an iPad-like tablet mounted to an arm. These are reports of nascent projects, which could go any number of ways, but as it stands, this sounds more in line with where the company’s robotic ambitions ought to be.

Home robots are coming, but when they arrive, they’ll still have a long way to go. That said, Jane won’t bring home Rosey until 2062, so we’ve got time.

Update: Amazon tells TechCrunch that reports of Astro's demise have been greatly exaggerated. The company notes that the robot is still available through its Day One program and "Amazon is still fully committed to our vision of bringing world-class consumer robotics solutions to the home and excited about the in-home experiences we’re inventing for Astro."