RobiDuo-B65: Bridging Autonomous Mobility and Bimanual Manipulation in a Single Robotics Platform

● Autonomous mobile robots have become increasingly capable over the past decade. Modern navigation algorithms can localize, map, and plan paths through complex environments with remarkable accuracy, allowing robots to move safely across factories, warehouses, laboratories, and public spaces. However, mobility alone rarely completes a task. A robot that can autonomously navigate to a workstation still requires the ability to perceive its surroundings, adjust its working height, manipulate objects, and interact with the physical world. This convergence of autonomous navigation and robotic manipulation has become one of the most active research areas in modern robotics.
● Building such a system from scratch is rarely straightforward. Researchers must integrate a mobile base, robotic manipulators, perception sensors, onboard computing, communication systems, and software frameworks before algorithm development can even begin. Hardware compatibility, mechanical integration, and system stability often consume far more development time than the robotics algorithms themselves.
● RobiDuo-B65 was designed to eliminate this complexity by providing a complete mobile manipulation platform for robotics research and industrial development.
● At the core of RobiDuo-B65 is a modular architecture that combines an autonomous mobile base, an electrically adjustable lifting column, dual robotic arms, onboard perception, and high-performance edge computing into a single research platform. Rather than limiting developers to a fixed hardware configuration, the system is designed to remain flexible throughout the development cycle. Different robotic arms, sensors, and computing modules can be integrated according to project requirements, allowing researchers to rapidly prototype new applications without redesigning the entire robot.
● Unlike conventional stationary manipulators, RobiDuo-B65 introduces an additional degree of flexibility through autonomous mobility. The robot is capable of navigating between workstations while carrying its manipulation system, significantly expanding its operational workspace. The integrated lifting mechanism further extends the robot's reach by allowing the upper body to adjust vertically, enabling interaction with work surfaces, storage shelves, laboratory equipment, or industrial fixtures at different heights. This combination of mobility, lifting, and dual-arm manipulation enables tasks that would otherwise require multiple independent robotic systems.
● Navigation remains the foundation of every autonomous mobile robot. RobiDuo-B65 supports multiple Robify mobile platforms, allowing developers to select the chassis that best matches their application requirements. Depending on the deployment scenario, the robot can be integrated with RobiHaul 1.0, RobiHaul 2.0, RobiX, RobiX-Nav, RobiS, RobiTrac Mini, or other Robify mobile platforms. This modular approach enables research teams to evaluate the same manipulation system across differential-drive, tracked, omnidirectional, or autonomous driving platforms while maintaining a consistent upper-body architecture.
● To support reliable autonomous navigation, the robot integrates multiple perception sensors including LiDAR, RGB-D cameras, IMU, and RTK positioning modules. These sensing technologies continuously generate environmental information for localization, mapping, obstacle detection, and path planning. Combined with ROS and Autoware compatibility, the platform provides a complete development environment for SLAM, autonomous navigation, multi-sensor fusion, and perception-driven robotics research.
● While mobility determines where a robot can operate, manipulation determines what it can accomplish. RobiDuo-B65 features a dual-arm architecture capable of performing coordinated manipulation tasks that are difficult or impossible for single-arm systems. The open mechanical interface allows different robotic arms to be installed according to payload, reach, and application requirements, providing developers with the flexibility to tailor the platform to research in industrial automation, human-robot collaboration, service robotics, or intelligent manipulation.
● The robot's vision system is mounted above the manipulators, providing a human-like observation perspective for object detection, pose estimation, visual servoing, and grasp planning. By combining visual perception with coordinated bimanual manipulation, RobiDuo-B65 supports advanced research in perception-guided manipulation, autonomous grasping, and AI-enabled robotic interaction.
● Supporting these perception and manipulation capabilities requires considerable onboard computing resources. RobiDuo-B65 integrates an industrial mini computer equipped with an Intel Core i7 processor and NVIDIA RTX 1650 GPU, providing sufficient computational performance for deep learning inference, visual perception, navigation algorithms, and robotic control. Instead of relying on external workstations, developers can execute complex robotics applications directly on the robot, significantly simplifying deployment and field testing.
● Equally important is the platform's open development philosophy. RobiDuo-B65 is fully compatible with ROS and Autoware while supporting Windows and Ubuntu operating systems. Standard communication interfaces and modular hardware architecture allow researchers to integrate additional sensors, robotic arms, industrial controllers, or custom software without modifying the robot's core architecture. Rather than defining a fixed workflow, the platform adapts to existing research pipelines and industrial development processes.
Remote development is just as important as onboard performance when deploying autonomous robots. RobiDuo-B65 is designed to simplify the entire development workflow by combining an expandable battery system with convenient remote access capabilities. Developers can remotely access the onboard computer through HDMI remote control solutions, allowing software development, debugging, algorithm testing, and system monitoring without remaining physically beside the robot. This enables researchers to iterate more efficiently, whether the robot is operating in a laboratory, warehouse, or remote testing environment. Combined with customizable battery configurations for different deployment requirements, RobiDuo-B65 supports extended experimentation while minimizing interruptions during long-term development and validation.
● The versatility of RobiDuo-B65 makes it suitable for a broad range of robotics applications. Universities can use the platform for courses involving autonomous navigation, mobile manipulation, perception, and artificial intelligence. Research laboratories can evaluate coordinated bimanual manipulation algorithms, visual servoing, and mobile manipulation strategies using a complete hardware system. Industrial developers can rapidly prototype warehouse automation, intelligent inspection, collaborative material handling, and flexible manufacturing solutions without investing months in custom hardware integration.
● As robotics continues moving toward increasingly autonomous and collaborative systems, future robots must do more than navigate—they must understand, manipulate, and interact with the world around them. Mobile manipulation is becoming a fundamental capability for next-generation service robots, industrial automation, and embodied AI research.
RobiDuo-B65 was developed with this vision in mind. By integrating autonomous mobility, adjustable working height, dual-arm manipulation, advanced perception, and an open robotics software ecosystem into a single modular platform, it enables researchers and developers to focus on innovation rather than system integration. Whether deployed in universities, robotics laboratories, or industrial R&D centers, RobiDuo-B65 provides a scalable foundation for the next generation of autonomous mobile manipulation systems.
.png)

























