Daimon Robotics Unveils Advanced Tactile Intelligence at IROS 2026

Daimon Robotics Showcases Pioneering Tactile Intelligence at IROS 2026

Daimon Robotics, a trailblazer in the realm of physical AI, is making waves at IROS 2026 by presenting its comprehensive tactile intelligence infrastructure. The theme of their presentation, 'From Touch to Intelligence,' embodies their mission to enhance robotic interactions with the physical world through advanced tactile perception.

The Need for Tactile Intelligence

In modern robotics, as machines are tasked with increasingly intricate activities, visual perception alone is no longer sufficient. Understanding physical interactions—such as contact force, material slip, and object response—is critical for successful robotic manipulation. Daimon Robotics is addressing this challenge by developing tactile sensors that provide vital information missing from traditional visual data. This integration allows robots to understand their physical environment more intuitively, equipping them for a wider range of tasks.

Building a Tactile Intelligence Ecosystem

At IROS, Daimon Robotics unveiled a multi-layered approach to tactile perception that consists of:

1. Perception: Featured in their innovative DM-Tac Series of vision-based tactile sensors, these advanced units capture over twelve tactile modalities through more than 110,000 sensing units. Live demonstrations showcased their capabilities in underwater sensing, durability tests, and the identification of empty packages—displaying their ability to maintain high-resolution perception even in challenging environments.

2. Data Integration: The Data-Nexus Acquisition System combines various data collection tools, networking facilities, and processing pipelines into a seamless data production workflow. This system consistently generates quality physical interaction data essential for training robotic models.

3. World Model: The Daimon-TWM model grounds intelligence on physical interaction data, enabling robots to perform real-time control and predictive decision-making for dexterous manipulation tasks.

Live Demonstrations of Daimon-TWM

Remarkably, IROS 2026 was the first venue for the public demonstration of the Daimon-TWM model since its debut. Attendees witnessed the robot performing two intricate autonomous manipulation tasks:

- Friendship Bracelet Stringing: This task showcased the robot's ability to pick up diminutive, engraved beads, align them precisely, and thread them onto a flexible string. The complexity required adaptations to subtle changes in physical interactions, allowing the robot to adjust its actions continuously instead of adhering to a predetermined sequence.

- Canvas Tote Heat-Transfer Printing: In this demonstration, the robot efficiently managed the entire heat transfer printing process from sticker placement to final adjustments. The robot utilized tactile feedback to enhance its responsiveness during various phases, ensuring precision in operations like peeling and placement which required adjustments based on the physical parameters of the materials involved.

The Future of Physical AI

Daimon Robotics aims to reshape how robots interact with their environments through its tactile intelligence infrastructure. By moving beyond mere visual recognition, they empower machines to learn from, adapt to, and perform complex interactions with the world around them. This transformation signifies a promising step toward a future where autonomous systems can mimic human-like capabilities in dexterous manipulation.

About Daimon Robotics

Established in 2023, Daimon Robotics emerged from the robotics research team at The Hong Kong University of Science and Technology. The company, headquartered in Shenzhen with a research and development branch in Hong Kong, focuses on creating a closed-loop physical AI ecosystem that integrates tactile sensing, data acquisition, and intelligent modeling. With their groundbreaking technology, they are committed to fostering robots equipped with true physical intelligence, capable of performing robust and dexterous tasks in the real world.