In May 2026, IO-AI Tech continued to advance the development of real-world data infrastructure for robotics and embodied AI.

As robots move beyond laboratories and into real-world environments, data is becoming more than a training resource—it is emerging as a foundational capability for the entire industry. Open developer tools, scalable data production, cross-disciplinary collaboration, and ecosystem-driven innovation are all becoming critical to the future of embodied intelligence.

Throughout the month, IO-AI Tech made progress across open-sourcing ROSView, global ecosystem participation, industry knowledge sharing, real-world deployment, academic collaboration, and standardization efforts—continuing to explore how high-quality data can help robots better understand and interact with the physical world.

Advancing the Developer Ecosystem with the Open-Sourcing of ROSView

In May, IO-AI Tech officially open-sourced ROSView, a high-performance robot data visualization platform designed for robotics and embodied AI development.

ROSView supports a wide range of commonly used robotics data formats, including MCAP, ROS 1 Bag, ROS 2 DB3, HDF5, and BVH. It provides rich visualization capabilities for images, point clouds, TF trees, joint states, and more, enabling developers to efficiently analyze, debug, and replay robot data.

Released under the MIT License, ROSView supports both commercial use and secondary development. It can be used as a standalone application or embedded into enterprise platforms and custom workflows. Through ROSView, IO-AI Tech aims to contribute to a more open and sustainable robotics developer ecosystem.

Learn More: https://io-ai.tech/en/news/open-source-rosview-embodied-ai-visualization/

Showcasing at BEYOND Expo 2026 and NVIDIA Inception Startup Events

In May, IO-AI Tech participated in the NVIDIA Inception Startup Showcase – Macau and exhibited at the NVIDIA Inception Startups Area during BEYOND Expo 2026 as a member of NVIDIA Inception.

At the event, we demonstrated the TeleXperience teleoperation system and the SenseXperience human data collection system. Highlights included one-to-many dexterous hand teleoperation enabled by exoskeleton gloves, as well as scalable human data collection workflows designed for real-world robotics applications.

As embodied AI continues to transition from research environments to practical deployment, high-quality data is becoming a key enabler of Physical AI. IO-AI Tech remains committed to advancing teleoperation, data collection, and data management technologies that support the next generation of intelligent machines.

Speaking at the Embodied AI Data Supply and Circulation Forum

On May 28, IO-AI Tech was invited to participate in the “Breaking Through & Rebuilding: Embodied AI Data Supply and Circulation Forum,” jointly organized by Chang Development and Link Interface Embodied Intelligence Center.

The event brought together leading organizations and innovators from across the embodied AI ecosystem, including ByteDance Volcano Engine, BOE, Schneider Electric, Haitian Ruisheng, and more than 60 AI and robotics companies, to discuss data generation, governance, circulation, and ecosystem collaboration.

During the forum, Gao Biao, Co-Founder and CTO of IO-AI Tech, delivered a keynote presentation titled:“From Operation to Data: Bridging the Last Mile of Embodied AI Deployment.”

The presentation shared IO-AI Tech’s experience in teleoperation systems, real-world data collection and large-scale data production, while exploring how high-quality data can accelerate embodied AI training and deployment.

Supporting Real-World Tea Harvesting with Teleoperation Technology

In May, humanoid robots participated in tea harvesting, transportation, drying, and processing tasks in the core white tea production region of Fuding, Fujian Province, as part of the first domestic energy relay event for the 2026 World Humanoid Robot Games.

During the project, IO-AI Tech's TeleXperience teleoperation system was integrated with the Inspire-Robots dexterous hand hand on the Unitree G1 humanoid robot. By capturing and mapping human motions in real time, the system enabled dexterous manipulation and data collection in a complex outdoor environment.

As embodied AI moves toward practical deployment, real-world operational data is becoming increasingly important. IO-AI Tech continues to explore how teleoperation and scalable data production can help robots learn from and adapt to the physical world.

SUN YAT-SEN UNIVERSITY Students Visit IO-AI Tech for Cross-Disciplinary Exchange

On May 13, faculty members and students from Sun Yat-sen University, City University of Hong Kong, and Xiamen University visited IO-AI Tech as part of an interdisciplinary seminar on Artificial Intelligence and Marxist Theory Frontiers.

Participants toured the company’s technology demonstration area and data collection facilities, experienced teleoperation-based data collection workflows, and engaged in discussions on embodied AI, data value, AI ethics, and human–robot interaction.

The visit provided a platform for dialogue between technological practice and humanities research, while exploring new approaches to interdisciplinary education in the age of artificial intelligence.

Exploring Industry Development and Standardization

In May, representatives from the Longnan Municipal Government and related departments visited IO-AI Tech for discussions on embodied AI industry development, standardization, innovation practices, and talent ecosystem development.

During the exchange, IO-AI Tech shared its experience in teleoperation, data collection, and real-world data infrastructure. Both parties discussed industry standards, data value creation, and ecosystem collaboration.

As embodied AI continues to grow rapidly, data collection centers and large-scale dataset initiatives are emerging across China. Standardized approaches to high-quality data production are increasingly becoming a critical foundation for the sustainable development of the industry.

Looking Ahead

From open-source tools to real-world deployment, from international ecosystem participation to academic collaboration, IO-AI Tech continues to advance the infrastructure that enables robots to better understand and operate in the physical world.

Our goal is to provide the leading real-world data infrastructure for robotics and embodied AI, partnering closely with customers and ecosystem partners to accelerate model development and real-world deployment.

We believe that enabling robots to operate effectively in the real world requires not only advanced hardware and intelligent models, but also scalable systems for collecting, managing, and utilizing high-quality real-world data.

Together with developers, researchers, and industry partners, we will continue working toward our vision of:

Bringing robots everywhere, for everyone.