Can a Chinese EV-maker reinvent itself as a robot firm?

At the ANNUAL tech showcase of Xpeng, a maker of electric vehicles (evs), the assembled crowd was surprised to see a female humanoid robot saunter along the catwalk. Speculation blew up online that the robot, known as iron, was so lifelike that there must be a human inside. The head of Xpeng, He Xiaopeng, decided, to the dismay of his robot team, to cut open its synthetic skin on stage the next day to reveal the mechanical components underneath. “People don’t dare to believe that such an advanced robot came from a Chinese startup,” Mr He told the crowd. “It’s like ten years ago, when many doubted that Chinese electric vehicles could be made well and go global.”

It turned out that Chinese firms could make evs well—too well and too cheaply for the comfort of foreign governments, which slapped tariffs on them. Now Mr He is revving up for the mass production of humanoid robots, flying cars and self-driving taxis after years of development.

He is bullish that he can achieve the same success. “By the time you and everyone else can see a trend, it’s already no longer your opportunity,” the baby-faced 48-year-old billionaire told The Economist at his headquarters in the southern city of Guangzhou. On April 1st Xpeng Motors changed its name to Xpeng Group to reflect its broader ambitions. “Think about Apple Computer when they changed their name to Apple Inc,” says Tu Le of Sino Auto Insights, a consultancy.

Can the darling of the ev industry really succeed with humanoids and flying cars? Mr He is not the best known of China’s tech bros, but he is one to watch. His early bet on giving evs tech features that now define the Chinese market—including a highly capable autonomous-driving system—paid off. His polymath projects have drawn parallels with Elon Musk’s.

Mr He studied computer science in Guangzhou in the 1990s, and says his first goal was to make money, which he did after co-founding UCWeb, a mobile internet browser, in 2004.  By 2012 UCWeb had 500m users. Alibaba, a tech giant, bought him out, making him a billionaire. Then he began fusing hardware and software in cars, convinced that they should be electric and smart. “It was a blind spot and an open window,” says Mr He, who co-founded Xpeng in 2014 and has run it since 2017.

His tech obsession led Volkswagen, the biggest foreign carmaker in China, to buy a 5% stake in Xpeng in 2023, for $700m. vw now uses Xpeng’s ai chips and driver-assistance systems in some of its Chinese models. Appearances of the diminutive Mr He beside the towering boss of vw China, Ralf Brandstätter, aptly contrast the nimble Chinese upstart with the giant legacy carmaker. “We are quite confident that he will come out as one of the winners,” says Mr Brandstätter.

Continuing to succeed in the cut-throat Chinese car industry will be hard enough; winning in the next round of innovation will be tougher still. Mr He will have to do three things. First, survive the “bloodbath” that he has long warned of in China’s ev market, which still has some 100 sizeable brands. Few ev firms make money as they undercut rivals for market share. Xpeng turned a profit for the first time only in the fourth quarter of 2025. While others cut r&d spending, Mr He says Xpeng plans to invest $1bn in ai this year. He wants to win by creating the best technology and then charging more for it. For reliable profits, Xpeng needs to sell more than the 429,000 evs it did last year and hopes to do so through boosting higher-margin exports.

The second issue will be technical.  Advancing from evs to flying cars and robots represents a screeching change. Many Chinese carmakers are moving in the same direction, but flying cars and robots are still early in their development. Mr He knows already how to bring ideas from one industry into another. And there is overlap in what is needed by evs and humanoids, for instance in motors, batteries, sensors and other components (Chinese ev suppliers already build parts for humanoids, including Tesla’s Optimus).

But the overlap should not be overstated, argues Poe Zhao, author of Hello China Tech, a newsletter. There is a substantial difference between building an ev and building a robot in areas such as balance and safety, he says, “A humanoid robot is not simply a car with legs.”

Third, the company must clear substantial commercial and regulatory hurdles. Mr He’s humanoids will enter mass production later this year, and flying cars in 2027. Xpeng already has 7,000 orders for its flying cars, but they still need more regulatory approval, supporting infrastructure and a viable business model. Boosters predict humanoids will be transformative. Morgan Stanley, a bank, reckons the market could be worth $5trn by 2050. But there are plenty of sceptics asking who will buy them and how firms will make money.

Turning cars electric and making them smarter is built on an established and growing market. There is no pre-existing demand for humanoids to build on (in China, the state is the biggest buyer). Robot-makers want humanoids to do laundry and cook and care for the elderly, but that still requires copious amounts of data collection and training. Analysts fear another bubble, as even state-owned carmakers pile into the humanoid hype.

Clearing all three of these hurdles will be much harder than the switch to making cars. Mr He has shown that he can dream big, and build popular products that others didn’t believe in. But the attempted transition will be his biggest gamble yet. Recent decades suggest that no one should bet against Chinese innovation and determination. But the pace of transformation in these sectors may be slower and less stunning, and the profits more distant. Mr He says he wants to deploy his first humanoid robots in Xpeng showrooms, already open in about 60 countries. Their job? To greet prospective buyers of his evs. ■