China could beat the world in the robotics frontier of technological competition – by solving the hardest problem in the industry – making human-like robotic hands.
The hand is one of the most complex parts of the human body. That complexity, comprised of nimble, nerve-filled appendages, has turned out to be a great problem in the robotics industry. Many parts of the human skeleton can be replicated in a robotic design with reasonable effort, but the hands have proved to be an insurmountable challenge.
Think about the seemingly casual tasks you complete with your fingers. Typing. Buttoning a shirt. Tucking in. Scratching your skin. Holding a pen. Sewing. Tying shoelaces. Those tasks take complex neurological software to execute. No machine in the history of industrialisation has ever been invented which executes such tasks as seamlessly as a hand.
Chinese companies have taken the front seat in leading innovations that could finally solve this problem, which is a key objective in the Artificial Intelligence (AI) race. If we are ever going to transition from the age of producing toy humanoid robots which are simply good at making gimmicks, dancing, and amusing children, to the age of very efficient, much more useful humanoid robots, we have to figure out the dexterity of hands, and manufacture humanoid robots with a capacity to complete tasks we humans easily accomplish with our hands.
China is deploying its manufacturing prowess in this area and is making great, inspiring progress. The incentives for this are numerous. Many jobs currently done by humans are often too dangerous, too dirty, too tiresome, too boring and uncreative. Daily tasks like cooking and folding laundry are not activities everybody enjoys, and many people would rather use their time for more creative work or simply for resting and entertainment. Even more critical, China and several developed countries are starting to witness challenges related to an ageing population and a shrinking workforce. All these are challenges whose answer could be in advancements in robotics, especially the manufacture of humanoid robots with nimble, dexterous hands.
The Chinese government is clear about its vision for multiplying the country’s economy. It has centred the building of ‘embodied AI’ in China’s development plans. In simple terms, embodied AI is the type of AI characterised by intelligent agents that are capable of interacting with the physical world through a physical body or agent.
As you read this, more than half of the factory robots built in the entire world each year are built in China. China is also leading in the deployment of automatons (self-operating machines that follow a predetermined sequence of operations). The challenge limiting larger-scale use of humanoid robots is still the limited abilities of their hands. Every robotics company in the world is grappling with how to make human-like robotic hands. Elon Musk, whose company Tesla makes the Optimus humanoid, has also usually highlighted the complexity of this challenge, remarking in 2025 that the ‘hand carries the majority of the engineering difficulty of the entire robot’.
According to Zhou Yong, the founder of China’s dextrous hand company LinkerBot, making a robotic hand is “one hundred times more difficult” than making a humanoid. He says: ‘Its dexterity is 10 times that of other body parts. But its volume is only one-tenth of the other body parts.
Yong’s company manufactures about 5,000 hands a month and has plans to double that figure as it chases a valuation of $6bn. He thinks that it would be much easier to excel at many human skills if we figure out the hand because human hands are the most important ability of human beings.
One of the advantages of figuring out this problem, which promises the highest utility to humans, is the ability to mass-produce cheap, nimble prosthetics for amputees.
The great advantage in competition that China boasts now comes from its large electric vehicle industry portfolio, which has many companies capable of producing robotic components at scale. It has advanced supply chains that make it cheaper to produce things like batteries and miniaturised motors, yet at the same time, it does so with peerless sophistication.
The United States, China’s nearest competitor, is heavily constrained by the supply chain in sourcing components. This makes it harder for its companies to compete against China. In 2025 alone, the percentage growth of robotic company registrations soared by 40%. The dextrous hands industry alone surpassed $7.4bn.
The challenge at stake now is making robots capable of manipulating tools. That’s what will make them more useful than simply moving. The problem in this industry is no longer locomotion but manipulation. The other challenge, then, is making software that can teach hands how to do things. It’s one thing to make nimble hands; it’s a whole other challenge to teach a robot to control them. This is what Chinese startups are working to solve at scale. They already have the vast amounts of data required to train spatial intelligence models, which can interact with the 3D world.
The writer is a senior research fellow, at the Development Watch Centre.
