AI robots have begun to go to nursing homes to accompany the elderly

Let's start with a change that you may not have noticed much, but is happening around you-robots are starting to go to nursing homes to accompany the elderly.

This is a scene in science fiction movies that is happening in 2026.

In a smart health care station in Yizhuang, Beijing, moxibustion robots are responsible for physical therapy, exoskeleton robots help the elderly with rehabilitation training, and a chess robot accompanies the elderly to play chess and make tea. In nursing homes in Shenzhen, robots can make round-the-clock rounds and take care of elderly people who need to be watched at all times at night. At a display point in Hangzhou, some robots can help bedridden elderly people turn over and monitor vital signs. Some can generate a TCM physical report in two minutes by looking at their tongues and simulating pulse measurement. Others can have emotional dialogue with the elderly to relieve loneliness.

2026 will be called by many as the "first year of large-scale implementation of elderly care robots."

When I first saw these news, my first reaction was that it was a little warm and complicated. What is warm is that technology is beginning to take care of these people who need to be taken care of most; what is complicated is that when we start to seriously discuss "letting robots accompany the elderly," there is actually a heavy reality that our generation cannot avoid.

In this article, I want to break this matter apart and talk about why it is now, how big the market is, and the most heart-warming question: Can robots really replace "companionship"?

!

First see clearly why it is now

The sudden explosion of an industry is never because technology is unilaterally mature, but because "demand" and "capabilities" collide at a certain point. AI elderly care is such a collision point.

First look at the demand side, put out a few numbers, and you will immediately understand why this matter is so urgent.

The number of elderly people over the age of 60 in my country has reached 323 million, accounting for 23% of the total population. It is expected that by around 2035, it will exceed 400 million, accounting for more than 30%, officially entering severe aging.

Of these 323 million people, about 35 million are disabled elderly people-that is, people who are no longer able to fully take care of themselves and need long-term care.

And what about the people who take care of them? The shortage of elderly caregivers in my country has reached 5.5 million.

If you put these three numbers together-323 million elderly people, 35 million disabled people, 5.5 million caregivers gap-a huge, structural "companionship and care gap" emerges.

What is more realistic is the change in family structure. The generation with fewer children and only children began to face the situation of "four elderly people, one couple, and one child." How can a young couple take care of four or more elderly people? It is expensive to hire nursing workers, and secondly, it is impossible to hire so many.

This is the real soil for AI elderly care-it's not that someone wants to play a gimmick. It is this society that is really short of manpower, and it is so short that new methods must be found.

Look at the ability side. In the past two years, physical intelligence, AI models, and robotic technology have moved forward together. Robots have changed from "toys that can flip after meetings in the laboratory" to things that "can work in real scenes." Being able to monitor, remind, provide simple care, and provide chats-these just meet the needs of elderly care.

The demand has been waiting for a long time, and the ability is just in place. So in 2026, this matter will explode.

How big is the market?

Big enough to exceed the imagination of many people.

The entire silver-haired economy is expected to exceed 30 trillion yuan by 2035. What concept is this? It is equivalent to recreating a super-large industry.

Specific to elderly care robots and AI companionship-some institutions predict that the global AI companionship market will soar from its current size to tens of billions or even hundreds of billions of dollars in 2030. The domestic market will be only about 1 billion yuan in 2024, but it is expected to grow several times by 2030, with a compound annual growth rate of 75%.

A compound annual growth rate of 75% is a very scary rate. This means that this market is turning upwards almost every year.

Capital and companies will certainly not miss it. A number of listed companies have already joined in-some product lines cover multiple types of exoskeleton rehabilitation, life care and companionship, and have been piloted in nursing institutions in many places; some have established joint ventures with robot companies like Ubisoft to specialize in making health care companion robots, which can provide the elderly with services such as companionship, reminder to take medicine, health monitoring, and emergency alarm.

There are also those who take the "bionic companionship" route-make them look like real animals such as pandas, long-haired cats, and puppies, with furry clothes to accompany the elderly to relieve boredom. These products are now mainly exported to sea, and are sold better overseas than domestically.

Policies are also being pushed hard. In June 2025, the Ministry of Industry and Information Technology and the Ministry of Civil Affairs jointly issued a pilot notice to specifically deploy research and scenario applications of intelligent elderly care service robots. The Ministry of Industry and Information Technology also released the "Humanoid Robot and Personalized Intelligence Standard System (2026 Edition)" to set rules and pave the way for the entire industry.

Demand, technology, capital, and policy, the four forces work together in one direction. This is what a typical "tuyere" should look like.

!

But pour cold water on it: It's not that easy

After talking so much hot things, I need to say something sober. Although elderly care robots have great prospects, if they really enter thousands of households, there are still three very hard obstacles in front of them.

The first hurdle is that technology is not reliable enough.

The scene of elderly care is completely different from moving boxes in a factory. The old man's body is fragile, and the robot helps him turn over, support, and take care of him. Any greater or lesser force may cause problems. At present, the industry generally recognizes that robots still have obvious shortcomings in force control accuracy and naturalness of movements. A robot that can throw things and is stiff. Do you dare to let it help your bedridden parents? This trust has not been fully earned by technology.

The second hurdle is who will pay.

This is the most practical question. A fully functional elderly care robot is not cheap, and the hardware profit of high-end accompanying robots can reach more than 60%, indicating that the pricing is very high. Who will pay for the money? Letting the elderly pay for it themselves is difficult for most families to bear and may not be able to afford it; letting their children pay for it is a considerable expense; if they expect elderly care institutions to purchase in bulk, they will also have to calculate the cost account. Market demand is rigid, but "demand" and "money to buy" are two different things. This payment chain has not yet been fully run.

The third hurdle and the deepest one-ethics and trust.

Giving the most vulnerable elderly people to a machine for the care of them has ethical weight in itself. Can robots make mistakes? Who is responsible if something goes wrong? How to protect the elderly's private data? More importantly-an old man being "taken care of" by a robot, can the robot give him the thing he lacks the most in his heart?

This leads to the most heart-breaking question.

robot can replace "care", but can it replace "companionship"?

I think this is the most important thing to think about in the whole matter.

Elderly care needs are actually divided into two levels. The first level is "care"-turning over, feeding, monitoring, reminding to take medicine, and being on duty at night. These are manual and standardized tasks. Robots are really capable of doing it, and they can do it more stably and tirelessly than humans. At this level, it is a good thing that robots replace humans and can truly reduce the burden on caregivers and children.

But there is another layer, which is "companionship"-the old man wants to find someone to talk to, the old man has someone who really cares about him when he is lonely, and the emotional connection of "being needed and remembered."

Robots can simulate companionship. It can talk to the elderly, play chess with them, and respond to you when you speak. Technically, it's becoming more and more like that.

But you and I both know that simulated companionship and real companionship are two different things. A robot saying,"I'm listening" is a completely different thing from your child actually sitting next to you listening to you nagging. The "emotional connection" that the old man wants is essentially "I am important to another person"-this is something that machines cannot currently provide.

So my judgment is this-the direction of elderly care robots is absolutely right and must be done, because the nursing gap is a real hard need, and robots can carry the "care" layer, which is a huge social value.

But we have to be clear: robots are here to fill the gap in "care", not to replace "companionship." If one day, we feel at ease not to visit our parents because we have elderly care robots and feel that "we have robots accompanying us anyway," that will be the biggest backlash from this technology.

Technology should be used to free up time and energy to accompany us, not to relieve us of the responsibility of companionship.

What does QKPFX4 mean to you

When it comes to you, there are a few real things--

First, if you have elderly people in your home who need care, elderly care robots are an option worth starting to pay attention to. Especially for those standardized care-monitoring vital signs, reminding them to take medicines, night duty, rehabilitation assistance-robots can really help in the past two years and greatly reduce the burden on you and your caregivers. You can pay attention to whether local elderly care institutions have introduced them, or pay attention to those health care companionship products that are already on sale.

Second, if you are considering career or entrepreneurial direction, the silver economy is one of the most certain directions in the next decade. A 30 trillion yuan plate, a 5.5 million caregiver gap, and an annual growth of 75%, these figures correspond to a large number of employment and entrepreneurial opportunities-not just making robots, but also elderly care services, aging-friendly transformation, rehabilitation care, silver entertainment, etc. This is an area that will continue to grow and has extremely rigid needs.

Third and most important-don't outsource companionship to machines. Technology can solve care, but not family. If your parents are still around, make more calls and go home more often. Elderly care robots can do many things for you, but they cannot replace the place of you as a "person" in their hearts. No matter how advanced the machine is, it cannot help you in this matter.

Last

Robots entering nursing homes to accompany the elderly are the rise of an emerging industry.

Digging down one layer, there is a gap of 323 million elderly people, 35 million disabled people, and 5.5 million caregivers. This huge structural gap forces the entire society to use new methods to fill it, and AI and robots are just catching up.

Digging further down is a reality that our generation has to face: as our parents grow old, we ourselves will eventually grow old, and fewer and fewer people can take care of us. Elderly care robots are essentially a technical answer given by this era to "provide security for the elderly."

This answer sheet can answer the question of "care" well. It can allow elderly people who are bedridden to be turned over in time, allow elderly people who live alone to be on duty at night, and prevent caregivers from being exhausted. This is remarkable progress.

But it couldn't answer the question of "companionship". The answer to that question has always only one word-person.

The robot can accompany the elderly to play chess, make tea and chat. But the person the old man is really waiting for is the one who will push the door open and call him parents.

No matter how powerful AI is, it cannot replace this matter. It should not be replaced.

!