IO-AI Tech Raises Hundreds of Millions of RMB to Accelerate Robotics and Embodied AI Deployment
IO-AI Tech, a robotics and embodied AI data infrastructure company, has completed a new financing round totaling hundreds of millions of RMB. The round was jointly backed by Shunwei Capital, Songhe Capital, and Shenzhen Capital Group (SCGC), alongside strategic investment from several leading robotics OEMs.
The new funding will support continued investment in core product development, team building, and global market expansion. IO-AI Tech will further strengthen its capabilities across robot teleoperation, human data collection, data management, and training workflows, with the goal of making embodied AI systems easier to develop, improve, and deploy in the physical world.
As robots enter the real world, data becomes critical to their continued improvement
Over the past several years, advances in foundation models have opened a new chapter for artificial intelligence, with embodied AI emerging as a major frontier.
As humanoid robots, mobile robots, industrial robots, and other robotic systems move from controlled environments into real-world applications, the challenge is evolving from simply “making robots move” to enabling them to understand their surroundings, perform increasingly complex tasks, and continuously improve.
At the center of this transition is data.
Unlike large language models, which can draw on vast amounts of text and visual content available online, robots need to learn through interaction with the physical world. Training robotic systems requires multimodal data that captures visual observations, human actions, robot states, and environmental feedback.
Much of this data cannot be obtained from the internet. It needs to be generated through interactions between robots, people, objects, and real-world environments.
This makes data infrastructure a critical component of the robotics stack. The industry needs systems that can efficiently collect, manage, process, and evaluate data, and feed the resulting insights back into model training and robot deployment.
IO-AI Tech is building this infrastructure to help bridge the gap between advances in robotics and embodied AI models and their deployment in the physical world.
IO-AI Tech: Connecting data, robots, models, and the physical world
Founded in 2023, IO-AI Tech is focused on building data infrastructure for robotics and embodied AI. Rather than focusing on a single part of the data workflow, IO-AI Tech has developed an integrated hardware and software platform spanning robot teleoperation, human data collection, and embodied AI data management.
By connecting data collection, management, training, and feedback into an integrated workflow, IO-AI Tech enables robotics companies, embodied AI model developers, and industrial customers to build and deploy robotic systems more efficiently.
IO-AI Tech currently offers three core product platforms:
TeleXperience
TeleXperience: General-purpose robot teleoperation and data collection system designed to provide a unified control interface across different robot platforms and configurations.
This enables human operators to transfer their expertise directly to robots, supporting robot training, remote operation, and real-world task data collection.
SenseXperience
SenseXperience: Real-world human data collection system comprising four core modules: Ego, Wrist, UMI, and Compute.
The platform captures egocentric, wrist-level, and hand-interaction data, providing scalable, high-quality training data for VLA, Physical AI, and World Model development.
EmbodiFlow
EmbodiFlow: Embodied AI data management platform that brings together data from multiple robots, sensors, and formats. It supports the full data workflow, from collection and annotation to management, review, and export, while connecting data pipelines with model training and deployment.
Together, the three platforms form an integrated infrastructure spanning data generation, data management, model training, and continuous feedback, allowing real-world data to drive ongoing improvements in robotic capabilities.
For robotics, data is not simply a one-time training resource. It is a continuously growing asset that becomes more valuable as robots interact with the physical world.
Every task a robot performs generates new data. That data can be processed, evaluated, and fed back into model training, enabling robots to take on increasingly complex tasks. Over time, this creates a continuous data flywheel for improving robotic intelligence.
Supporting the global robotics ecosystem
Today, IO-AI Tech serves more than 150 customers worldwide, including robotics OEMs, embodied AI model companies, and industrial customers, and has adapted its hardware to more than 80 robot systems. These include humanoid robots, robotic arms, dexterous hands, and other robot systems.
The company is also working to foster a more open ecosystem for robotics and embodied AI. IO-AI Tech has open-sourced multiple real-world datasets and embodied AI data tools, including LeRobot Studio and ROSView, and plays an active role in international robotics competitions, research collaborations, and industry initiatives.
By making data and development tools more accessible, IO-AI Tech aims to support greater collaboration and accelerate innovation across the global robotics ecosystem.
Looking ahead: Bringing robots everywhere, for everyone
As robotics moves toward large-scale deployment, IO-AI Tech will continue to invest in the infrastructure needed to make robotic systems more capable, adaptable, and effective in real-world environments.
IO-AI Tech will further develop its capabilities in general-purpose robot teleoperation, human data collection, embodied AI data management, and remote intervention, while expanding partnerships across the global robotics ecosystem.
By working with robotics companies, model developers, and industry partners, IO-AI Tech aims to accelerate the adoption of robotics across industries and help bring capable robotic systems into everyday life.