The first time I thought about robots in the context of caregiving was nearly 6 years ago, when I listened to an episode of Caterina Fake’s podcast, Should This Exist, that explored a robot caregiver being piloted in a Washington, D.C., retirement home.
Back then, I felt the slightest twinge of an ick about the thought of delegating caregiving — even just the social interactions — to a robot.
Today, the ick is gone.
Robots that could take care of us as we got older were thought to be so far into the future that it was pushing the edge. Now, the edge is here.
What changed?
The models improved, algorithms got more advanced, and other types of ambient solutions were successfully integrated into care pathways.
But most importantly, the pain got more acute.
We are all getting older, and there are fewer and fewer human caregivers waiting for us.
In 2025, the number of adults in the U.S. over the age of 50 grew to outnumber children under 15 years old. By 2050, we are estimated to have a global population of 1.5 billion people aged 65 or older.
At the same time, caregiving as a profession (the CNA who comes in to help turn your grandfather so he doesn’t get bedsores or the aide who comes in to make sure mom took her heart medication) has been chronically underpaid and undervalued.
Caregiver pay has not kept pace with demand. Burnout is high. The workforce pipeline is thin.
Our healthcare and caregiving infrastructure is already crumbling and was not built to absorb the coming wave.
While we are living longer, healthier lives, and as longevity science promises to extend healthspan even further, the care needs of our population will continue to increase as everyone ages.
We simply do not have enough human caregivers to meet those needs.
Demand for Care is Growing Faster Than the People Who Can Provide It
60% of older people want to grow old living in their own homes. But as they get older, less mobile, and less able to live independently, 24/7 in-home care is often too expensive and inaccessible for those who need it. Even when financially feasible, families face a growing shortage of home care workers.
Within care facilities, nearly every U.S. nursing home has reported facing staffing shortages. This crisis is forcing senior care facilities to close their doors, with two-thirds of nursing homes sounding the alarm that they may not survive another year.
By 2034, the U.S. care sector is projected to face 6.1 million unfilled jobs. Roughly 35% of these vacancies will be caused by workers leaving caregiving as a profession, while more than half will result from people exiting the labor force entirely.
Given the conditions, it’s no surprise. According to the U.S. Bureau of Labor Statistics, the median annual salary for direct care workers, home health aides, residential care aides, and nursing home assistants is $26,000 to $39,530.
There are ways to address the human labor problem. Increasing caregiver pay, creating new workforce pipelines, introducing new models for financing long-term care, and even creating a dedicated caregiving immigration visa are all potential fixes that are vital to explore.
But we’re trying to plug a hole that isn’t a trickle, it’s a tsunami.
Physical AI has the potential to step into that gap.
Companion Robots
Caregiving robots first showed up as a solution to the ‘loneliness epidemic’.
ElliQ by Intuition Robotics launched in 2022. This countertop robot with a tablet interface and a bright, happy, head-shaped light mounted on a minimalist base is designed to make living as a senior alone at home easier and more enjoyable, contributing to overall health outcomes by tracking mood and pain, reminding them of the day ahead, and keeping them in touch with family and friends. Intuition Robotics reports users interacting with ElliQ 30 or more times a day and 94% of users reported an improvement in their quality of life.
A newer entrant in the companion robotics space, Andromeda Robotics, launched “Abi”, the ‘happiness assistant’ robot, and placed it into 22 aged care homes in Australia. They’ve since expanded into the U.S. and are focused on deploying their ‘happiness assistants’ to replace one billion hours of loneliness in aging care and assisted living facilities.
Watching Andromeda Robotics set up at West Tech Fest in Perth, Australia (2024)
Abi’s buyer is nursing homes and facilities, while ElliQ has created distribution through government contracts and is targeted at older people aging in isolation and seniors being cared for by family members. The goal of both robots is similar: reducing loneliness and providing support for the things that aren’t as easy any longer: capturing day-to-day health information, staying in touch with loved ones, and remembering important details.
Physical and mobility assistance
This is the biggest opportunity for physical AI in the caregiving space. The shortage of nurses also means a shortage of physical labor. Lifting, turning, and toileting assistance are all vital to caregiving, but place physical strain on nursing staff and create workers’ compensation liabilities for care facilities.
Twolabs.ai, a recent Y Combinator company, is building humanoid robots that assist nursing homes and senior living centers in caring for the elderly, starting with feeding, dressing, medication, daily tasks, and companionship.
This already frees up a huge amount of labor, and there is also a future where the robots can help with toileting, lifting, and other more dexterous tasks.
A group of MIT engineers developed the Elderly Bodily Assistance Robot, or E-BAR, an eldercare robot designed to address the challenge of helping older people sit, stand, and prevent falls. Designed for use in the home, the E-BAR comes with inflating airbags on the robot’s arms that can catch a person before they fall.
Japan’s AIREC (AI-Driven Robot for Embrace and Care) is a humanoid prototype developed as part of the JST Moonshot Program. This family of robots has two current iterations, both aimed at helping older people with bodily functions and daily care needs like showering, toileting, diaper changing, and mobility, to reduce pressure-induced bedsores.
The opportunity for these types of assistance robots goes beyond care for the elderly. Hospitals are also a future use case which would broaden their utility and TAM to general healthcare.
Cognitive monitoring
Nearly half of all dementia cases worldwide can be traced to ‘modifiable risk factors’, many of which are within our control. Engaging in cognitively stimulating activities and carefully managing chronic conditions can have a meaningful impact on long-term brain health.
When family can’t be there every day to monitor an aging person’s cognitive health, AI-powered cognitive care can. Robots can use real data to track cognitive decline over time so that we become less reliant on the subjective monitoring that family and caregivers typically do.
HOMY Robotics is the creator of Pepper, an AI in-home robotic caregiver, to solve this problem (among others) for family caregivers. One of Pepper’s core functions is to provide gentle memory prompts, daily conversation, and interesting news to support healthy cognition while monitoring for signs of decline.
Can America Compete in the Global Market for Caregiving Robotics?
The race to win the robotics market might become one of the defining technology battles of the next decade. South Korea, Singapore, and Japan have, in one way or another, outpaced the U.S. over the years. But today, China, our biggest competitor, is winning on the factors that matter most: price, scale, and speed.
China became the leader in industrial robotics, creating a $47 billion market in 2024, and deploying more than 2 million robots in manufacturing settings in 2025, more than any other country, and a 6X increase from the year before.
Morgan Stanley estimates that China’s humanoid robot market will reach $2 billion in 2026 and $15 billion by 2030. Top-down policy and government support catalyzed the growth of China’s robotics market, and the care robot market will grow on the heels of its humanoid market.
The Beijing Economic-Technological Development Area is one of several major hubs where local government, private care agencies, and state-backed robotics firms are working together to rapidly scale the deployment of next-generation caregiving robots nationwide.
In 2025, Tiantai Robotics announced a joint venture to produce 10,000 humanoid robots for eldercare and in-home assistance. This will be the largest deployment of assistive robots in any country to date.
Unitree Robotics, China’s humanoid robot company known for training their robots to do backflips and martial arts, released an AI model that allows for more complex home care tasks to advance general care assistance for elder care.
Unitree’s willingness to open source has been an accelerant. Other robotics companies are building on the UnifoLM-WBT dataset and a South Korean joint venture released their “Gen P” humanoid built on the Unitree’s G1 platform with dexterous hands that help aging adults complete precision tasks.
China’s sophisticated supply chain, America’s limited manufacturing capabilities, and tariffs are the risk soup for American robotics companies.
The U.S. leads on software, technical sophistication, and AI integration. But hardware is hard, and you can’t iterate software advantages into a manufacturing base that doesn’t exist.
Unitree and Agibot shipped roughly 23 times more units last year than U.S. robotics companies Tesla, FigureAI, and Agility Robotics combined.
In September 2025, the U.S. Commerce Department launched a Section 232 investigation into the imports of robotics and industrial machinery, framing China’s state-backed robotics industry as a national security threat. Chinese robot dogs and humanoids already face steep tariffs, and the administration is signaling that more restrictions may be coming. The White House also convened a roundtable with executives from Boston Dynamics, SpaceX, Siemens and others to map out how to rebuild domestic robotics and components manufacturing from the ground up.
You can tax the finished product, but you can’t tax your way into a supply chain that took China decades to build.
The Caregiver Shortage is No Longer a Future Problem
Human capacity is the biggest bottleneck in care. But robots are quickly becoming more dexterous and capable of becoming true partners to human caregivers in healthcare settings.
The call for robot caregivers is not new. What is new is the increased pain in the form of more older people, a shrinking pool of caregivers, and the maturation of the technology that finally makes it viable.
If caregiving robots can alleviate the caregiving pressure on a stretched professional workforce and help our parents and grandparents age in place, we have the chance to redefine health outcomes for generations.