Kuri

» see heykuri.com as it lived, on the wayback machine

~ the obit ~

i missed kuri. wrong region for the marketing push, wrong price for impulse purchase, wrong category for any consumer base i recognized. mayfield robotics shipped almost entirely inside the united states at $699 plus tax. honestly the home robot category in turkey in 2017 mostly meant a robot vacuum, the small irobot roomba clones that had finally hit altinmarka and teknosa shelves around that year. a wheeled companion with no arms and no vacuum was not the home robot turkish buyers were looking at.

mayfield robotics was a bosch internal venture. anyway, the company spun out of the bosch startup platform around 2015 in redwood city california, the team had a few ex-pixar designers on the industrial side which is why kuri ended up with the wall-e silhouette. unveiled at ces 2017 in early january, the demo was the showroom hit of the consumer-robot category that year, the press wrote it up as the cute counterweight to jibo's more anthropomorphic design. for whatever reason the unit itself was wheels, two eye cameras, a body speaker, and an array of microphones, no manipulator arms, no spoken-language output beyond a vocabulary of burbles.

kuri started shipping to backers in december 2017. the early reviews mostly liked the hardware quality and were lukewarm about what the robot actually did once it was in the house. for whatever reason the burbling-only voice was supposed to be a feature for households with small kids, the navigation worked, the cameras streamed back to a phone app. the gap was that the robot did not really do anything beyond wander and photograph. you could tell it to find a person and it would, you could not tell it to do almost anything else.

~ the rap sheet ~

~ how it landed ~

the ces 2017 reception was the kuri high point. tech press named it one of the best home products of the show, the unit photographed well, the pixar design language read as approachable in a category that had been dominated by uncanny humanoids. for whatever reason the launch wave assumed the household-robot market was a near-term consumer category, the same assumption powered jibo a year earlier and pepper the year before that on the japanese side. anyway, the consumer market did not actually exist at that price band, the household robot was still in the buy-it-because-you-write-about-tech tier in 2017.

the bosch wrapper around mayfield was the unusual thing. an industrial conglomerate had bankrolled a redwood-city startup to test whether a consumer robot could become a product line on the consumer-appliance side. the answer ran out in mid-2018 when bosch concluded there was no business fit. for whatever reason that conclusion arrived faster than most of the other consumer-robot flops of the era because mayfield reported up to bosch quarterly and the conglomerate did not have patience for a long discovery curve.

~ the cracks ~

july 25 2018, mayfield announced it was pausing production of kuri after the first shipping batch. the language in the press release was that the company was halting the business to focus on next steps, which in startup-press code meant a search for a buyer. for whatever reason no buyer materialized. august 21 the company announced full closure, october 31 was the last day of operations. backers who had paid for kuri but not yet received units were refunded in full, which was the unusually clean part of the wind-down. most consumer hardware flops of this era left preorder customers stranded, mayfield did not.

but the postmortems from techcrunch and the verge from august 2018 mostly named the same diagnosis. the home robot value proposition in 2018 was not yet a thing households would buy. anyway, jibo had hit the same wall a few months earlier, the smart-display market was eating the friendly-home-device niche from below at a much lower price point, alexa and google home were doing the voice work that a home robot was supposed to do without needing a chassis (or whatever it was called). honestly the kuri burble was charming and the device was well built, neither of those facts answered the basic question of what the unit was for inside a household.

~ what survived ~

kuri sits in the home-robot graveyard next to jibo and the earlier sony aibo run. anyway, the broader pattern across that whole cluster was the same one, well-funded consumer robots that demoed cleanly and shipped at a price point too high for the actual utility once the unit was in a normal house. the bosch wrapper around mayfield meant kuri got buried faster than the comparable startups but the underlying gap was identical.

in '26 the heykuri.com domain has been recycled. mayfield robotics is a wikipedia stub. the shipped units became collector items in robotics-hobbyist circles, a few open-source projects exist that try to bring the robot back online without the dead cloud service. honestly that is the same long tail every consumer robot from this window inherits, a small community of owners reverse-engineering the firmware after the company that built the cloud half of the product has gone (close enough).

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »