[waterheater]

I worked at iRobot some years back. I'll tell you exactly why iRobot is struggling: iRobot, at its core, built robots.

Does that seem contradictory? Consider that their innovations, what they're known for, lies in robotic morphology. iRobot's Cool Stuff Museum in their HQ has some groundbreaking robots which still are awesome (like a ocean buoy robot powered by the waves). They cut their teeth doing research grants and developing military robots. Their ethos was rooted in service; when 9/11 happened, they packed up a prototype robot and drove to Ground Zero to use that robot to search for survivors in the rubble. Still, their military robots were controlled by humans, and the original Roombas turned at random angles to fully clean a room.

Around 8 years ago, iRobot spun off their military and commercial robot divisions into independent companies (Endeavor Robotics and Ava Robotics, respectively; Endeavor was eventually acquired by FLIR) to focus exclusively on the consumer market. The consumer robotics industry was shifting, so iRobot needed to change. They bought Evolution Robotics to acquire computer vision IP and started doing research in that area. Advanced software capabilities were now important, so they shipped vSLAM (monocular SLAM representing state using pose graphs; they've published quite a few papers on this), persistent mapping, automatic room segmentation, and much more. They kept researching new consumer robots, including a home security robot which didn't make it out of the lab.

The thing is, consumer robots will ALWAYS be a luxury item. Nobody NEEDS a Roomba, and the iRobot folks understood that reality. So, iRobot leaned into that and worked to make the Roomba the premium option: a refined industrial design language with their main products appealed better to the high-end market, and they released an autoevac Roomba. I'm guessing that their primary market segment saturated, so they figured they'd release new physical products, but that puts them in direct competition with Dyson (who utterly failed to produce a good robot vacuum cleaner, mind you), so good luck there. I don't believe iRobot ever resorted to selling their customer data, which I'm sure Chinese robotics companies regularly do to increase revenue.

I also think iRobot's failure to launch their robot lawnmower really hurt them, probably hurting morale than anything else. Tons of people worked on that project, and the technology stack was genuinely impressive for seven years ago (using UWB beacons placed in the yard for localization and mapping). UWB was still an emerging technology in 2017 (still a few years before AirTags released). It makes me sad they couldn't ship it in the US, since the technology was better than virtually all other robot lawnmowers on the market.

iRobot manufactures their robots in China, but if I remember correctly, that's exclusively because the Chinese restrict exports on key electrical components like resistors. There was a desire to onshore manufacturing, but because the Chinese cornered the market, it was basically impossible to make the costs work out.

The bottom line is, the robot vacuum market is basically a solved problem, and there's a very good reason that floor cleaning robots continue to be the only viable consumer robot product.

I have fond memories of those good and kind iRoboteers. I wish iRobot all the best as they move forward and try to find a well-constrained everyday problem which can be solved by a low-cost robot.