At first glance, it looks like a routine cleaning job. However, a wheeled robot then moves beside a cleaner, Lei Xiaoli (43), then stretches out its mechanical arm to collect the scattered rubbish.
For Lei, this isn't about replacing a worker, but rather the beginning of a new working partnership. She's learning to share tasks with a machine designed to navigate the chaos and unpredictability of a real home.
The robot, mounted on a wheeled base and equipped with two mechanical arms, is part of a home cleaning service launched in March by Shenzhen-based startup X Square Robot through a partnership with home services platform 58.com.
At 149 yuan (1 yuan = Rp2,536) for a three-hour session, the service divides tasks between humans and machines. The robot handles basic household chores like picking up trash, tidying shoes and toys, and changing trash bags, while the cleaner focuses on more detailed tasks in the kitchen, bathroom, and bedroom, as well as interacting with clients.
"When I first saw it, I was very curious and kept observing how it worked," Lei said. After more than 20 sessions working together, Lei said he's gotten used to the collaboration, and the robot has become increasingly proficient.
"The robot can identify areas that need cleaning on its own, thus reducing my workload," he said.
Some early customers in Shenzhen were quick to try out the new service.
A customer, Ge Landong (pseudonym), deliberately scattered household rubbish in his living room to test the robot's capabilities.
"Whether it's a banana peel or a small chicken bone, the robot can accurately recognize it, pick it up, and then throw it in the trash," he said. "Its perception and precision surpassed my expectations."
However, there are still some limitations. Ge said the robot's relatively large size prevents it from reaching tight spaces and is only capable of handling simple tasks, meaning human cleaners still need to follow up. An engineer is also present during service visits to ensure everything runs smoothly.
X Square Robot acknowledged these limitations in a social media post on April 15, stating that the robot currently moves slower than humans and still needs refinement to adapt to complex home environments.
The company said it also launched the service in Beijing on April 21 and plans to expand to more cities, with a target of deploying 1,000 units by 2026.
The trials demonstrate a silent evolution underway: in the era of tangible artificial intelligence ( AI), robots are not just demonstrating their capabilities, but are beginning to learn to work in everyday life scenarios.
"The home environment is much more complex and dynamic than a factory. In a sense, the home is the best testing ground for robots," said Liu Shaoshan, an embodied AI expert at the Shenzhen Institute of Artificial Intelligence and Robotics for Society .
Liu said China has advantages in hardware supply chains, rapid product iteration, and a diversity of real-world application scenarios, factors that allow companies to refine their systems more quickly through real-world feedback.
Beyond household cleaning, robots could play a role in elderly care, providing daily companionship and support in a sector that has long been hampered by labor shortages, he said.
Positioned as a future industry in China's latest five-year plan, tangible AI is predicted to receive a new growth boost.
According to a Morgan Stanley report, China's robotics market is projected to grow from US$47 billion (US$1 = Rp17,324) in 2024 to US$108 billion in 2028, with service robots registering a compound annual growth rate of 25 percent.
Across China, provincial-level regions are increasingly pushing to develop this sector. In Guangdong Province, service robot output is projected to rise 11.2 percent by 2025, accounting for approximately 80 percent of the national total. In Beijing, production figures surged 47.6 percent year- on-year (yoy), reflecting the sector's strong momentum.
Industry players say that as robots increasingly take over routine tasks, aspects that remain for humans, such as judgment, adaptability and interaction, may become more central to the nature of work.
How that balance evolves could shape not only the future of cleaning jobs, but also how humans and machines coexist in everyday life.
For Lei, the changes are already starting to have an impact in the real world.
"I hope the robot can learn to clean windows or carry heavy objects," he said. "That would reduce the physical burden and safety risks for us."
