The service industry is a vital pillar of social welfare, with elderly care services being a particularly critical component. By the end of 2025, China's population aged 60 and above is projected to reach 320 million, and this demographic is expected to exceed 400 million by 2035. According to the "2025 Research Report on the Current State of the Elderly Care Worker Profession," China is expected to face a shortage of over 5 million elderly care workers in the next five years. As various robots designed for meal delivery, rehabilitation, and companionship enter the realm of elderly care, the question arises: to what extent can they actually share the burden of caregiving services?
In Wuzhen, a picturesque water town with an aging population rate exceeding 30%, robots are already taking on labor-intensive tasks. At the Wuzhen Smart Elderly Care Center during lunch hour, meal delivery robots steadily transport food boxes to the doors of rooms housing seniors with limited mobility along pre-set paths. In the activity room, a group of silver-haired seniors follow the lead of a robot named "Xiao Bai" for exercise sessions; while their movements may be out of sync, the room is filled with laughter and joy. "The robots deliver meals and medicine to us, replacing a lot of the work normally done by care aides. They are very efficient and convenient," remarked Pan Jiancheng, a resident at the center.
"Bathing seniors with limited mobility previously required two people working together. It was not only physically demanding but also carried various risks," explained Zhong Yueying, Director of the Wuzhen Smart Elderly Care Center. "Now, with portable bathing machines, one person can handle the task. The elderly don't even need to leave their beds, which greatly improves their comfort. The time and energy saved allows us to spend more time talking with them." Tu Chaorong, Vice President of Shenzhen Zuoke Technology Co., Ltd., added, "To address the challenge of bathing bedridden seniors, we developed a portable bathing machine that frees caregivers from heavy physical labor, allowing them to devote more energy to warm, human communication."
This substitution of physical labor represents the initial phase of robot integration into nursing homes. Currently, caregiving robots primarily handle tasks that are either difficult or impractical for humans, such as round-the-clock safety monitoring, precisely repetitive rehabilitation exercises, and physically strenuous lifting and transferring. Chen Diansheng, a professor at the Robotics Institute of Beijing University of Aeronautics and Astronautics, believes that robots in elderly care are currently in a stage of human-robot collaborative assistance, where they take over some auxiliary care tasks. Looking ahead, the advent of safe and controllable general-purpose humanoid robots holds the potential for single machines to perform multiple functions, offering versatile services. "Smart elderly care will continue to evolve towards greater intelligence, personalization, comprehensiveness, and community integration," noted Huang Jianyuan, a professor at the School of Public Administration at Hohai University.
However, the real-world challenges are far more complex than imagined. Integrating robots into care services presents significant difficulties. "A simple task like pouring water involves unexpected complexity: the robot must see the cup, understand the command, anticipate the water volume, coordinate its manipulator, and apply just the right amount of force to avoid spilling," explained Zhu Songchun, Dean of the Beijing Institute for General Artificial Intelligence. "Home environments are vastly more complicated than laboratory settings, and daily needs vary immensely from person to person."
An incident at a nursing home in Shenzhen illustrates one challenge: a transfer and lifting robot, due to a center-of-gravity design that did not conform to ergonomic principles, caused noticeable discomfort to an elderly user during operation. Zhao Yong, a member of the National Technical Committee for Robotics Standardization, analyzed that robots struggle to provide the meticulous care that human workers can offer to seniors with limited mobility, primarily due to insufficient capabilities in gentle manipulation. Actions like turning a person over, feeding, or assisting with walking require a soft touch. Robots currently lack sensitive tactile feedback, and improper force control can cause discomfort or even safety risks for the elderly, failing to match the subtlety of human care.
"This is fundamentally a process of identifying 'genuine, rigid demands'," said Tong Fei, founder of Zhongcheng Yuansheng Robotics. "Elderly care scenarios are highly fragmented. General-purpose robots can't adapt to all needs. Companies must customize solutions for each specific requirement – how to navigate multiple floors, how to interact with seniors who speak slowly, how to communicate with those experiencing hearing loss. This is refined through painstaking, case-by-case effort."
In 2025, a team led by Professor Jiang Hanqing from Westlake University unveiled the Mawarm flexible variable-stiffness robotic arm and the HEART series of robots. This robotic arm maintains rigidity during routine operations to ensure precise grasping and stable movement. However, upon collision, its internal programmable pneumatic system inflates within milliseconds, softening to absorb impact. This enables rapid transitions between rigidity and flexibility, enhancing safety and adaptability in care environments.
Beyond achieving "softness," an even greater challenge is making robots conform to the unique physical characteristics of each individual senior. "Everyone has different height, weight, gait, and even muscle shape. Patients often have additional individual differences like abnormal muscle tone or joint limitations. Creating a product that fits everyone is extremely difficult," explained Wang Tian, founder of Hangzhou Cheng Tian Technology Development Co., Ltd. The company leverages vast amounts of clinical data to build its own human engineering database and biomechanical simulation models, which simulate human-robot interaction during device use. To date, the company has accumulated clinical application data from over 900,000 uses. Based on this data, exoskeleton robots can more accurately recognize the movement intentions of different seniors, thereby providing appropriately tailored power assistance for users with varying body shapes, gaits, and muscle tone abnormalities.
The warmth of technology often lies in the details. Many nursing homes report issues such as robots failing to understand local dialects or displaying on-screen prompts that are too technical. After a few failed attempts, seniors often become reluctant to use the technology again. Research from a team led by Jin Yong'ai, a professor at the School of Population and Health at Renmin University of China, indicates that 82.9% of middle-aged and older internet users face various obstacles online, with this proportion rising to 87.9% among those aged 70 and above. Jia Junfeng, an associate professor at Capital Medical University, cautions that a robot's emotional interaction is based on programmed settings; it cannot comprehend complex emotions, raising the possibility of "objectifying" people.
Sun Bo, CEO of Beijing Hai Baichuan Technology, shared a touching detail during an interview: their Xuanjing robot supports six dialects, including Sichuanese and Cantonese. Initially, the team viewed this merely as a feature. However, during testing at a nursing home, an elderly person with a strong Sichuan accent spoke to the robot in their local dialect. When the robot responded accurately, the senior's eyes lit up immediately. "He found it very difficult to speak Mandarin and was afraid of being laughed at for speaking poorly, but using his dialect put him at ease," Sun Bo said. Test data showed a dialect interaction accuracy rate of around 95%, but more important than the number was the significant increase in the senior's willingness to communicate. Ironically, a joke feature initially designed to amuse the elderly had relatively little effect.
"The integration of robots into homes will undoubtedly be a gradual process," Zhu Songchun stated. Tong Fei expressed that an aging population should not be viewed as a societal liability but rather as a touchstone for "technology for good." When robots can understand the dialects of the elderly and comprehend their needs, technology can truly contribute to the transition from simply "being cared for" to "enjoying later life." "This represents a unique advantage for China in the field of artificial intelligence," Zhu Songchun added.
Amid the current wave of AI enthusiasm, provinces and cities across China are actively developing local application scenarios to promote the deep integration of artificial intelligence with all sectors of the economy. In December 2025, the Ministry of Civil Affairs and seven other departments jointly issued the "Several Measures on Cultivating Business Entities in Elderly Services to Promote the Silver Economy," encouraging the development of the elderly care service robot industry. The measures aim to address the needs of families and institutions for daily nursing, emotional companionship, and social support for the elderly, fostering cross-industry collaboration and technological integration among robotics, medical rehabilitation, and smart home technologies to provide comprehensive, intelligent support for seniors.
From a familiar local accent to a gentle, supportive touch, the warmth of technology is embedded in these details: in understanding dialects, in soft physical contact, and in fostering a genuine sense among the elderly that "someone is thinking of them, someone is caring for them."
Comments