Gasgoo Munich- Embodied intelligence startup INFIFORCE has closed nearly RMB 1 billion in combined Series A and A+ funding, the company announced on August 14.
The round was led by Dunhong Asset alongside major state-owned platforms, with participation from Zhejiang University Sci-Tech Innovation Group, Yandu State-owned Investment Control, and Lishui State-owned Assets. Existing shareholder Creation Ventures (CCV) also increased its stake, as industrial capital and professional institutions collectively back the race to build the robot "brain."
Image Source: INFIFORCE
The capital will be deployed across three key areas: R&D for the AtomBrain embodied intelligence and causal world model; iterative upgrades to the full-stack AI infrastructure, DataGrid; and large-scale delivery verification of multi-form robots in real industrial scenarios. Backed by successive funding rounds, INFIFORCE is securing the capital needed to transition its technology from the laboratory to the factory floor.
A critical industry bottleneck is emerging: many embodied models rely heavily on third-person demonstration data. While robots can mimic these actions, they struggle to interpret their environment from a first-person perspective or predict the consequences of their movements. Once they leave the controlled demo environment, operational stability drops sharply. To address this, INFIFORCE has focused on an Ego-Centric data strategy since 2025.
Unlike traditional approaches, Ego-native data captures first-person interaction experiences from actual human operations, simultaneously recording visual feeds, action trajectories, spatial relationships, and environmental feedback. This method does more than just switch the camera angle; it leverages daily human activities to generate training material at scale. Compared to physical robot collection, this approach promises to significantly lower the cost of acquiring high-quality data while providing robots with a richer bandwidth of real-world experience.
Plans call for INFIFORCE to soon launch the world's first embodied model built on high-quality Ego data, completing a technical closed loop that connects Ego-native data, world models, and the AtomBrain embodied brain. The industry's competitive center of gravity is shifting from robot hardware to the challenge of enabling continuous experience acquisition and autonomous evolution. INFIFORCE's approach aims to answer a core question for the sector: how can robots continue to enhance their capabilities after entering the physical world?
To date, INFIFORCE has deployed and verified multi-form robots across hundreds of industrial scenarios in dozens of Chinese cities. With this new funding in place, the company plans to accelerate the commercialization of its technical loop, pushing embodied intelligence beyond the demo stage to meet real-world demands in industrial and logistics sectors.
