DexForce uses virtual data training to cut humanoid robot costs in China

Traders Union
All News
Financial News
DexForce cuts robot costs in China

Highlights

DexForce raised 1 billion yuan ($147 million) in Series B funding at a valuation exceeding 10 billion yuan, backed by Lenovo Capital and Lens Technology.
DexForce uses synthetic data from virtual simulations to train humanoid robots, reducing reliance on costly and time-intensive human-generated real-world datasets.
DexForce's robots and software are deployed commercially by Toyota, Samsung, Foxconn, and others, operating in factories and serving customers in over 15 Chinese cities.

Simulation strategy and funding backdrop

As reported by Reuters, DexForce is betting that synthetic data generated in virtual environments can address one of the industry's main bottlenecks, training humanoid robots to perform a broad range of tasks without human assistance. The company creates digital versions of workplaces or specific jobs, trains robots inside those simulations and then transfers the learning to physical machines.

Chief Executive Jia Kui says collecting only real-world data is too inefficient for the scale needed to reach what he describes as "physical artificial general intelligence," a stage where a humanoid can complete new tasks on command in unfamiliar settings. He says human trainers may need to repeat a movement more than 100 times to produce precise examples, making simulation a faster and lower-cost alternative.

Founded in 2021 by Jia, who is also a machine-learning professor at the Chinese University of Hong Kong, Shenzhen, DexForce employs about 500 people. The startup is backed by Lenovo Capital, Lens Technology and several state-owned funds, and recently raised 1 billion yuan, or $147 million, in a Series B round at a valuation of more than 10 billion yuan.

Commercial adoption and industry implications

DexForce's software is already used in industrial settings by Chinese and foreign manufacturers including Toyota, Samsung, Panasonic, Foxconn, Haier and Hisense. Its humanoid robots also make coffee and serve customers at malls and tourist sites in more than 15 cities across China.

That commercial push comes as many Chinese robot makers combine simulated and real-world data while still describing access to high-quality training data as a major constraint. Dozens of humanoid robot data training centres are being built nationwide, often at high cost and with heavy government subsidy support.

China has developed humanoids that can outperform humans in running and jumping, but the harder business test remains whether they can reliably perform useful tasks in working and consumer environments. DexForce's approach highlights a broader industry effort to lower training costs and speed deployment as competition in the sector intensifies.

Our earlier report on AI mega-rounds described how venture capital is becoming increasingly concentrated in fewer, larger funding deals as AI start-ups absorb a growing share of global fundraising. We noted that this shift pushes investors to write bigger cheques for a smaller set of companies, intensifying competition for top AI bets and raising concentration risk as overall deal counts fall.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.