
At the Smart Factory + Automation World 2025 held this March at COEX in Seoul Korea, Mintrobot presented its groundbreaking approach to robotic system integration under the theme "Next-Generation Robot Automation System Based on SDR (Software Defined Robotics)."

Traditional robotic automation processes follow a linear and fragmented structure: mechanical design, electrical design, assembly, robot and vision programming, and final PLC integration. However, these systems often face repeated cycles of redesign, reassembly, and reprogramming due to failed test runs. This leads to significant project delays, increased development costs, and high employee turnover rates driven by unsustainable workloads.

To address these pain points, Mintrobot introduced SDR as a revolutionary solution. SDR structures the system as a set of modular components, allowing the system to operate flexibly through simple configuration adjustments without the need to modify the development modules.
Unlike the conventional PLC-centered approach, where multiple robots and vision systems are tangled together during integration, the software-based SDR approach utilizes a middleware layer. This layer simplifies robot and vision integration by providing a unified communication channel.
Mintrobot’s proprietary robotics middleware enables seamless control of a wide range of robot brands—such as Universal Robots, Rainbow Robotics, and even various Chinese-manufactured robots—through a single, consistent API. Moreover, it allows the integration of both high-end and cost-effective vision systems under the same unified framework.
A key benefit of SDR is hardware independence. Systems can now be built and operated using software alone, providing consistent robot application management regardless of hardware variations. This facilitates easy adoption of the latest technologies, such as AI and deep learning, which are inherently software-driven, enabling the realization of highly sustainable robotic integration systems.
Additionally, SDR opens doors to advanced virtual environments. Mintrobot emphasized that controlling robots via software means that the same control logic can be applied to virtual robots within digital twin environments. Echoing NVIDIA CEO Jensen Huang’s comments on Physical AI, Mintrobot highlighted how virtual simulations can dramatically reduce the costs associated with building physical testbeds by enabling realistic testing and training in virtual environments.

Mintrobot showcased successful SDR applications in real-world scenarios, including food-tech automation proposals for Samsung Welstory and lab automation solutions for Seegene, both demonstrated within virtual environments using its proprietary virtual robot controllers.
The company is actively developing multi-robot applications that integrate mobile robots, robotic arms, grippers, and vision systems into mobile manipulators. Mintrobot is also preparing for a future where quadruped robots and upper-body humanoid robots will work in collaborative ecosystems—further validating the importance of SDR-driven integration.
Mintrobot’s SDR platform is powered by three core technologies:
- Robotics Middleware – Provides unified API control for various off-the-shelf robots.
- Universal Robotics Controller – A versatile control unit capable of managing various types of robotic arms and interfacing with middleware.
- Virtual Robotics Controller – A virtualized controller for managing robots within digital twin environments.
Notably, Mintrobot’s Universal Motion Controller has enabled collision detection on single-axis robots—an industry-first—by emulating physical phenomena in virtual environments. This represents Mintrobot’s next-generation, reverse digital twin technology that goes beyond conventional digital twins focused solely on data feedback.

A Mintrobot CEO remarked, "Adopting an SDR-based system is not just a system upgrade; it is the key to reducing costs, shortening development cycles, improving maintainability, preventing talent attrition, and enhancing your organization’s technological capabilities.”
The presentation concluded with a thought-provoking question for the audience:
"Is your robot system software-defined?"
Author: Minsk Kang