RobotEra Is Not Waiting for the IPO to Prove Its Robots Work
RobotEra Is Not Waiting for the IPO to Prove Its Robots Work
On the floor of a logistics sorting center somewhere in eastern China, a robot is picking a jumbled parcel off a conveyor belt, reading its label, and dropping it into the correct bin. It has been doing this, without a break, for nineteen hours straight. The facility has eleven more of them.
That is the image Beijing-based RobotEra wants investors and customers to carry away from its latest funding announcement. The company said this week it has closed a new financing round of more than $200 million, led by logistics giant SF Group, with participation from HSG, IDG Capital, Hillhouse Investment, CICC Capital, and a long list of industrial partners including Geely, BAIC, Lenovo, Haier, and funds affiliated with China Unicom and ICBC. The round brings RobotEra's total capital raised in just two months to nearly $350 million, following a RMB 1 billion strategic round in March that already pushed its valuation past $1.4 billion.
The scale of the funding is remarkable. The more interesting question is what SF Group thinks it is buying.
SF Group is one of China's largest express delivery operators. It runs hundreds of sorting facilities, employs tens of thousands of couriers, and faces the same long-term labor cost pressures that every logistics company in the world is trying to solve. RobotEra's L7 humanoid robot is currently deployed in more than ten of SF Group's logistics centers across north, east, and southern China, alongside facilities run by China Post. In the second quarter of 2026, RobotEra says it began delivering units at a scale of thousands.
That makes SF Group both RobotEra's biggest deployment customer and the lead investor in its latest funding round. China Post, the state postal service, occupies a similar position. These are the companies most directly exposed to the labor cost equation that humanoid robots are supposed to transform. They are also the ones writing the checks.
This is not a pattern unique to RobotEra. Geely and BAIC are automotive manufacturers with enormous workforces and active factory automation programs. ICBC Capital is the investment arm of the world's largest bank. The investor roster does not look like a group of financial funds doing due diligence on a startup. It looks like a coalition of potential customers who have decided the best way to influence the technology that may displace them is to own a piece of it.
RobotEra calls this strategy private rounds over IPOs, distinguishing itself from competitors Unitree and AgiBot, both of which are moving toward public listings. The logic is straightforward enough: an IPO requires polished financials, quarterly disclosures, and the discipline of a public market narrative. A private round with industrial partners does not. What it provides instead is deployment volume, validation signal, and information. SF Group's investment tells other potential customers something about how the robots are performing inside SF's own facilities. That signal is worth more to a logistics company considering a purchase than any press release.
The technical foundation is real enough to invite scrutiny. RobotEra's ERA-42 architecture is a Vision-Language-Action model that processes raw visual input and generates motor controls in real time, allowing a robot to handle the high variance of e-commerce goods without constant reprogramming. The company says it develops more than 95% of its core components in-house, including the XHand, a 12-degree-of-freedom dexterous manipulator it claims is the first full direct-drive design of its kind in the industry. The roster of hardware adopters on the client side includes Boston Dynamics, NVIDIA, and Apple, which suggests RobotEra's components have value beyond its own robots — though the nature of those relationships is not disclosed.
RobotEra says its robots have achieved 85% of human-level efficiency in logistics environments while maintaining continuous 24/7 operation. Those are the company's own figures, not yet audited by an independent party. Shenzhen, Huzhou, Hangzhou, Hefei, and Beijing appear in earlier disclosures as deployment sites. The company says it expanded to ten facilities in Q1 and Q2 2026.
The tension in this story is not whether humanoid robots can sort packages. The evidence from RobotEra's deployment data, whatever its limitations, suggests they can do it at a scale that was not commercially viable twelve months ago. The tension is what the investment pattern actually signals. Incumbents funding the technology that may displace them is either a sophisticated hedge against labor cost inflation or a way of controlling the pace and terms of their own disruption. A logistics company that owns equity in the robot maker has different incentives around adoption timelines than one that does not.
RobotEra has chosen to stay private and accumulate strategic investors rather than pursue the validation and liquidity an IPO provides. SF Group has chosen to lead a round in the company whose robots run in its own facilities. These are not the decisions of parties that are uncertain about the direction of the technology. They are the decisions of parties that are certain — and want to be positioned on both sides of it.