Aidin Robotics to Build AI-Integrated Grippers for Doosan Cobots

Recap

Source: 테크월드, report of Sept. 30, 2026

According to the report, Aidin Robotics has been selected as a joint research and development partner for the 'K-On-Device AI Semiconductor Technology Development Project,' a national initiative led by the Ministry of Trade, Industry and Energy. The company will collaborate with Doosan Robotics to develop sensor-integrated grippers for Doosan's next-generation collaborative robots. The project is scheduled to run until December 2030, with Doosan Robotics overseeing the overall effort and developing the robot body and controller, while Mobilint is responsible for the AI semiconductor component.

Aidin Robotics will focus on the development of both 2-finger and multi-finger grippers. The company plans to embed its proprietary force/torque and tactile sensors directly into the gripper design from the outset. This integration is intended to allow the grippers to automatically adjust force based on the object being handled, facilitating safe collaboration with humans. The safety test results from these grippers are also expected to support Doosan Robotics' compliance with international safety certifications.

Additionally, Aidin Robotics will develop a 'Force/Tactile Data Collection SDK.' This tool will record and replay force and tactile data in a standard format, providing a foundation for robot AI training data. The developed grippers are planned to undergo field validation in logistics and manufacturing environments before commercialization. The two companies previously signed a memorandum of understanding for physical AI co-development in November of the previous year, and this project marks an expansion of that cooperation into product development and commercialization.

Context

This project is part of a broader South Korean government effort to build a domestic ecosystem for physical AI, specifically focusing on on-device AI semiconductors. By integrating locally developed AI chips, collaborative robots, and grippers, the initiative aims to reduce reliance on foreign components in the robotics supply chain. Collaborative robots, or cobots, are designed to work alongside humans without safety cages, making force and tactile sensing critical for safe interaction. The inclusion of a data collection SDK highlights the growing importance of high-quality sensory data for training AI models that control physical actions.

Robot's take

The integration of force and tactile sensors directly into the gripper design is a significant step toward more dexterous and safe human-robot collaboration. By embedding these sensors at the design stage rather than adding them as afterthoughts, Aidin Robotics may achieve more precise force control and better data quality for AI training. The development of a standardized SDK for force and tactile data could be particularly valuable for the robotics community, as high-quality sensory data remains a bottleneck for training physical AI models. However, the success of this project will depend on the practical performance of the grippers in real-world logistics and manufacturing environments, and the ability to seamlessly integrate the domestic AI semiconductor with the robot's control system. It remains to be seen whether the resulting system will offer a competitive advantage over existing solutions that may use more established foreign components.