Nursing-home operator Charm Care puts ¥300M into Donut Robotics’ cinnamon humanoid

Charm Care Corporation, a Tokyo Stock Exchange Prime-listed operator of more than 100 private nursing homes in the Kansai and Greater Tokyo regions, will invest ¥300 million (about $2 million) in Donut Robotics by subscribing to newly issued shares, under a capital and business alliance. The deal was announced in Japan on August 31 and in English on October 5, 2026.

The money is tied to a deployment. Donut Robotics’ cinnamon humanoid has been on trial at a Charm Care home since July 2026, mainly holding conversations with residents.

What does the robot do in the home today?

Mostly talk. Since July the trial has centred on conversation with residents, according to Humanoids Daily’s account of the two companies’ announcements. Donut says patrol functions are developed and being improved; DealStreetAsia describes cinnamon 1 as using a vision-language-action model to patrol facilities, recognise its surroundings and describe what it sees. Responding to nurse calls, when a resident presses a call button, is a development goal rather than a current capability.

Neither company has reported any measured reduction in staff workload, or how much of the robot’s activity is autonomous versus remotely operated. The investment is also not a commitment to put cinnamon in all of Charm Care’s homes; it funds wider deployment of physical AI across them, starting from one site.

Why a care operator is investing

For Charm Care the stake buys influence over a product it wants to use. For Donut, a customer with more than 100 sites is a route from demonstration to repeat deployment, which is where most humanoid makers stall. Donut plans to move toward Japanese production, starting from overseas hardware and adding its own chip modifications and assembly in Japan.

Japan’s humanoid activity is increasingly organised around specific service jobs: a station guide on the Osaka Metro, GMO’s robot-swapping “humanoid ambulance” and Panasonic’s 2029 production goal. Elder care is the largest of those jobs and the hardest to automate, because the work is physical and the people being cared for are frail.