Positioning general-purpose picking robots, Yincheng Intelligence raises tens of millions of US dollars in financing

On May 25, it was reported that the general-purpose picking robot developer Yincheng Intelligence officially announced the completion of a multi-million dollar angel round of financing.

This round of financing was led by Detong Capital, with an industrial capital partner participating, and Yunxiu Capital acting as exclusive financial advisor. As the concept of embodied intelligence continues to gain momentum, this startup founded in 2023 is attempting to transfer the technical logic of autonomous driving to industrial and logistics sorting scenarios.

Public information shows that founder Pu Qunyan graduated from Fudan University and Carnegie Mellon University, previously worked at Google, and later joined Silicon Valley autonomous driving unicorn Nuro as founding team technical leader, with over 15 years of experience in AI algorithms and Robotaxi deployment.

Transitioning from autonomous driving on open roads to logistics robots in restricted scenarios, the underlying logic is technological reuse and dimensionality reduction. The core of autonomous driving lies in handling complex, chaotic interactions in the physical world; Yincheng Intelligence is trying to apply such abilities—based on visual perception, autonomous planning, and control—to similarly disorganized and non-standardized sorting operations in express logistics scenarios.

The main pain point in current logistics sorting is the variety of parcel types, inconsistent packaging specifications, and massive SKU volumes. Traditional industrial automation equipment mostly relies on teaching programming, is instruction-driven, and struggles to adapt to highly disordered and dynamic work environments, leading to a mismatch between supply and demand.

The core of Yincheng Intelligence’s business is its independently developed “NECESSI” embodied intelligence end-to-end large model. This solution aims to enable robots to “understand” different item shapes and autonomously decide grabbing strategies, just as autonomous vehicles identify road conditions, freeing them from heavy reliance on preset programs.

Based on a closed-loop of data from actual operations, the system can continuously collect long-tail data for autonomous learning, theoretically giving the equipment greater scenario generalization capability.

In the B-end robot market, technological advancement must ultimately be tested by efficiency and cost. According to official data released by Yincheng Intelligence, its sorting system currently has a picking success rate of up to 99.99%, with single item operation time controlled at about 2 seconds, benchmarking and even aiming to surpass traditional devices in terms of efficiency.

For commercialization, the company chose the express sorting market, with a potential scale of more than 20 billion RMB, as its first battle scenario. Currently, its picking robots have entered the business workflows of top logistics companies such as SF Holding, JD Group, and China Post.

It is reported that the company has recently signed batch orders, with cumulative order sizes reaching several tens of millions RMB. This data shows that Yincheng Intelligence has initially completed the transition from laboratory technical verification to real-world commercial closed loop, and will soon start mass production in 2026.

After completing this round of financing, Yincheng Intelligence plans to focus its capital input on mass production operations in 2026, new generation model R&D iterations, and seek expansion into multi-industry scenarios. Looking ahead to 2027, the company plans to extend its application boundaries from logistics and express delivery to e-commerce, pharmaceuticals, food & beverage, and automotive parts, aiming for revenue to reach the 100 million RMB level.

Objectively speaking, Yincheng Intelligence's development potential and challenges coexist:

The potential lies in the high growth of technology. The company can migrate its mature Robotaxi algorithm architecture to indoor sorting scenarios with relatively higher fault tolerance and closed environments, providing a feasible and imaginative commercial path. Once the large model data flywheel is operational, the marginal cost of cross-industry replication will drop significantly.

The challenge lies in B-end engineering delivery and generalization barriers. The hardware track tests not only the intelligence of algorithms, but also supply chain stability, manufacturing cost control, and on-site engineering delivery capabilities.

Additionally, as the business expands from high-fault tolerance express logistics to pharmaceutical or precision manufacturing fields with strict compliance and accuracy requirements, whether its end-to-end model can maintain equally high generalization and stability will still require rigorous market testing.

Against the backdrop of capital favoring embodied intelligence, Yincheng Intelligence has showcased impressive technical background and business expansion capability at the outset. Going forward, how to turn this first-mover technical momentum into solid engineering and cost barriers will be the core test on its path to scaling up rapidly.

Risk Warning and DisclaimerThe market comes with risks, and investments need to be made cautiously. This article does not constitute personal investment advice, and has not considered individual users’ specific investment goals, financial situation, or needs. Users should consider whether any opinions, viewpoints, or conclusions in this article are suitable for their particular circumstances. Investment based on this article is at your own risk.