한국어
ArchiveCase2017–2024, later unknown

Labrador Retriever: a home robot that carries things, silent since 2024

Labrador Retriever is a robot that carries items around the homes of people with limited mobility. It was unveiled at CES 2022 and taken up by nine care organisations, but the company has posted no news since January 2024.

Labrador® Retriever Quick Overview · Labrador Systems (YouTube)
MakerLabrador Systems (Calabasas, California, USA). Founded in 2017 by Mike Dooley (CEO) and Nikolai Romanov (CTO)
UnveiledCES, January 2022. Home pilots from February 2021
Price$250 refundable deposit. $1,499–1,500 upfront plus $99–150 a month for 36 months (sources differ). Contiguous 48 US states only
FormA wheeled robot with two shelves. Footprint 18 × 23 in. Upper deck adjustable from 25 to 38 in; carries up to 25 lb. Retriever lifts trays; Caddie has fixed shelves
SettingHomes and senior living. People with limited mobility or chronic pain
ReachNine first business customers for Retriever Pro in November 2022. Unit numbers not disclosed
NowWebsite still up, but the CES 2022 reservation programme has closed. Last company news in January 2024
Ended byNot confirmed (whether the company is still active is unknown)
Website up; no news since 2024; general sale unconfirmed

Timeline

  1. 2017Company foundedLabrador Systems
  2. 2021.02Home pilot
  3. 2022.01CES debutPre-orders opened
  4. 2022.11Pro adoptedRetriever Pro, 9 organizations
  5. 2024.01News stopsNothing found after

A robot to carry things for hands holding a cane

Labrador Retriever is a home assistive robot made by Labrador Systems of Calabasas, California. Mike Dooley and Nikolai Romanov founded the company in 2017, and Dooley had a hand in developing the Roomba. The robot is a wheeled unit with two shelves. Its upper deck adjusts from 25 to 38 inches, and it carries up to 25 lb. There were two models: Retriever, which can lift trays, and Caddie, with fixed shelves. Called up by touchscreen, app or Alexa voice command, it finds its way with 3D vision. Dooley said that once you start using a cane or walker, your hands become your extra pair of legs.

Unveiled at CES 2022, then nine business customers

The company began home pilots in February 2021 and unveiled the robot at CES in January 2022. The Amazon Alexa Fund and iRobot Ventures took part in its funding. Reservations took a $250 deposit, then $1,499–1,500 upfront and $99–150 a month for 36 months (sources differ). The company aimed for full production in the second half of 2023. It ran a multi-state pilot with Nationwide in March 2022, and in November announced nine first business customers for Retriever Pro. They included care and senior-living organisations such as On Lok PACE and Masonic Homes of California. At the time TechCrunch reported that the home Retriever was still in development and testing. The robot was demonstrated with an Echo Show at CES 2023, and the last item on the company’s press page is a CES panel in January 2024.

“Quite possibly the most exciting home robot”

IEEE Spectrum’s Evan Ackerman wrote that “this is quite possibly the most exciting home robot I've seen in years.” He reported that two pilot users asked whether they could invest in the company. Dooley said the 2021 pilots showed a deep need for practical help with activities at home. He also said the burden on caregivers was growing at an unsustainable rate.

A moving shelf instead of a human shape

Labrador took on a single job, carrying things, with shelves and trays rather than arms or a face. That contrasts with NEO, which aims to handle housework in a human shape. Pilot users and experts reacted well, but whether home sales continued after 2024 is unconfirmed.

How long can a robot that does one thing well stay useful in the home?

What we haven’t confirmed

  • Whether the home Retriever reached full production and general sale
  • Whether the company is still operating (the site is up, but there has been no news since January 2024)
  • How many robots were deployed
  • The monthly fee ($99, $100, $149 or $150, depending on the source and model)