ICRA 2026｜TESOLLO DG-5F and MANUS Metagloves Pro Haptic Demonstrate Next-Generation Robotic Teleoperation for Embodied AI
- 1 day ago
- 4 min read
ICRA 2026 Highlights: The Rise of Dexterous Manipulation and Embodied AI
As humanoid robots and Embodied AI continue to advance, one of the key challenges in robotics is enabling machines to understand human movements, learn complex skills, and perform precise manipulation tasks in real-world environments.
At ICRA 2026 (IEEE International Conference on Robotics and Automation), one of the world's leading robotics conferences, researchers and industry leaders showcased the latest breakthroughs in humanoid robotics, robot learning, and intelligent manipulation.
Among these developments, dexterous manipulation, robot teleoperation, and human-to-robot skill transfer became major areas of interest, highlighting the importance of creating more natural interfaces between humans and robots.
At the event, TESOLLO demonstrated its DG-5F five-finger dexterous robotic hand integrated with MANUS Metagloves Pro Haptic, showcasing a robotic teleoperation system that combines hand tracking, robotic control, and haptic feedback technologies.
From Automation to Embodied AI: Why Dexterous Manipulation Matters
Traditional industrial robots have primarily focused on repetitive and structured tasks, such as assembly, welding, and material handling.
However, with advances in artificial intelligence, sensing technologies, and computing power, the next generation of robots is moving toward Physical AI — intelligent systems that can perceive, reason, and interact with the physical world.
This evolution is driving the development of Embodied AI, where AI models are connected with robotic bodies, sensors, and environments to learn and execute real-world tasks.
Human hands represent one of the most advanced examples of dexterous manipulation. The ability to grasp, rotate, adjust, and interact with objects enables humans to perform highly complex tasks.
For robots to operate effectively in homes, factories, and service environments, similar capabilities are required. This is why dexterous hands, motion capture, and haptic feedback technologies have become essential components in modern robotic development.
TESOLLO DG-5F: Enabling Human-Like Robotic Manipulation
The TESOLLO DG-5F is a five-finger dexterous robotic hand designed for humanoid robots, AI robotics research, and advanced robotic applications.
Unlike conventional two-finger grippers, a dexterous robotic hand provides multi-finger coordination and more human-like manipulation capabilities, allowing robots to perform more complex interactions with objects.
Potential applications include:
Humanoid robotics
Robot Learning
Imitation Learning
Robotic Teleoperation
AI robotics research
Intelligent manufacturing
By improving robotic hand dexterity, robots can move beyond fixed automation tasks and develop the ability to adapt to more complex environments.
MANUS Metagloves Pro Haptic: Connecting Human Motion with Robotic Control
During the ICRA 2026 demonstration, TESOLLO DG-5F was integrated with MANUS Metagloves Pro Haptic, creating an intuitive interface for human-to-robot interaction.
MANUS Metagloves Pro Haptic is a high-precision hand tracking and haptic feedback glove designed to capture detailed finger movements and provide immersive interaction capabilities.
Through real-time motion capture, the operator's hand movements can be transferred directly to the TESOLLO DG-5F robotic hand, enabling precise control of individual finger movements.
With haptic feedback technology, operators can receive physical interaction information from the robotic system, improving control accuracy and making remote manipulation more natural.
How Robotic Teleoperation Accelerates Robot Learning
For modern AI robots, high-quality training data is one of the most important factors for improving intelligence.
However, generating sufficient real-world manipulation data remains a major challenge.
Robotic Teleoperation provides an effective approach by allowing humans to
demonstrate tasks directly:
Human operators perform manipulation tasks through motion capture devices.
The robotic system reproduces the demonstrated movements.
Interaction data is collected from real-world operations.
AI models use this data for Robot Learning and skill improvement.
These human demonstrations provide valuable datasets for Imitation Learning, helping robots acquire complex manipulation skills more efficiently.
As a result, teleoperation has become a critical technology pathway for advancing humanoid robots and Embodied AI.
TESOLLO × MANUS: Building an Integrated Robotics Development Pipeline
The ICRA 2026 demonstration reflects a broader trend in robotics development.
Future robotics innovation will not depend only on hardware performance, but on the integration of multiple technologies:
Technology | Role |
Dexterous Robotic Hand | Enables precise manipulation capabilities |
Motion Capture | Captures human demonstration data |
Haptic Feedback | Improves human-robot interaction |
AI Models | Learns and optimizes robotic skills |
ROS 2 Integration | Provides flexible robot control architecture |
Simulation Platforms | Accelerates development and testing |
Together, these technologies create a complete development pipeline for next-generation AI robots.
Axis3D: Integrated Solutions for AI Robotics and Motion Capture
Axis3D is the official Taiwan distributor of TESOLLO, providing advanced robotics and motion capture solutions for enterprises, research institutions, and educational organizations.
Our solutions include:
TESOLLO dexterous robotic hand integration
MANUS hand tracking and haptic glove solutions
Xsens full-body motion capture systems
Humanoid robot development environments
ROS 2 system integration
Unity and Unreal Engine applications
Robot Learning and Embodied AI development workflows
By connecting hardware, software, and AI development processes, Axis3D helps organizations explore new possibilities in intelligent robotics.
Conclusion: The Next Stage of Robotics After ICRA 2026
ICRA 2026 demonstrates that the future of robotics is moving beyond hardware capabilities toward more intelligent, adaptive, and human-centered systems.
The integration of TESOLLO DG-5F and MANUS Metagloves Pro Haptic highlights how dexterous hands, motion capture, and AI learning technologies can work together to advance robotic teleoperation and Embodied AI.
As robots become capable of learning from human demonstrations, new applications in manufacturing, research, and real-world environments will continue to emerge.
FAQ
What is ICRA 2026?
ICRA (IEEE International Conference on Robotics and Automation) is one of the world's leading robotics conferences, showcasing the latest research and industrial innovations in robotics and automation.
What is TESOLLO DG-5F used for?
TESOLLO DG-5F is a five-finger dexterous robotic hand designed for humanoid robots, Robot Learning, robotic teleoperation, and advanced robotics research.
How does MANUS Metagloves Pro Haptic support robotic teleoperation?
MANUS Metagloves Pro Haptic captures precise hand movements and provides haptic feedback, allowing operators to control robotic systems more naturally and collect high-quality demonstration data.
Why is teleoperation important for Embodied AI?
Teleoperation enables robots to learn from human demonstrations, providing valuable real-world data for Robot Learning and Imitation Learning.
