MEVION:

Abstract

The global competition for developing robotic foundation models is intensifying, and dual-arm robot data collection systems have become important platforms for manipulation research. ALOHA is a representative low-cost open-source system, but its limited force and speed make it difficult to handle heavy objects or perform fast manipulation. To address this, we developed MEVION, a low-cost open-source dual-arm robot data collection system capable of generating greater force and speed. All parts can be sourced through e-commerce, and by extensively utilizing sheet metal welding, its large body structure is constructed with a small number of components at low cost while simplifying assembly. MEVION is equipped with four 6-DoF arms with parallel grippers. Each arm weighs 7.0 kg and has a maximum torque of 60 Nm, and the entire system can be constructed for about USD 14,000. The elbow joint adopts a closed-link mechanism similar to those used in quadruped robots, reducing distal mass and enabling higher force and speed output at the end-effector. We demonstrate that MEVION enables data collection for object manipulation tasks not previously possible and supports imitation learning-based motion generation.

Design Overview

MEVION is primarily divided into arm and hand components. Its essential metallic structural parts comprise 11 components for the arm and 6 for the hand, totaling 17 components. When joint angle limiters and cover parts are included, the total number of metallic components is 21. The arm has six degrees of freedom, and the hand adopts a slider-crank parallel gripper mechanism where a single motor drives two fingers.

Metal Parts and Sheet Metal Welding

MEVION reduces part count through extensive use of sheet metal welding. The Shoulder-Link and Lower2-Link of the arm, as well as the Hand-Slider-Base of the hand, integrate large and complex shapes into single units via sheet metal welding, while the remaining parts are manufactured through machining. All metallic components are designed to be compatible with MISUMI's meviy online machining and fabrication service, enabling the ordering process to be completed through e-commerce.

Performance Compared with Existing Systems

Compared with existing open-source leader-follower dual-arm data collection systems, MEVION is a tabletop system with all-metal arms and hands, 6+1 DoF per arm, 7.0 kg arm weight, 0.83 m arm length, 60 Nm maximum torque, 20.4 rad/s maximum speed, and about USD 3,500 cost per arm. Compared with ALOHA, MEVION is about 1.6 times heavier while achieving approximately three times the maximum torque and seven times the maximum speed. Its peak payload at maximum extension is 7.6 kg.

Software

MEVION is controlled through a unified Python script, enabling both MuJoCo-based simulation and real-world control to be executed with the same code. It supports real-time visualization of the robot state in RViz via ROS and interactive rendering before sending control commands using Scikit-Robot. MEVION operates on either 24 V or 48 V power, adopts CAN communication, and uses software-based gravity compensation rather than a dedicated mechanical gravity-compensation structure.

Experiments

We evaluated MEVION through teleoperation and imitation learning experiments. Teleoperation tasks included bottle cap opening, object packing, frying pan operation, 3.6 kg dumbbell manipulation, and Daruma Otoshi. Imitation learning tasks included towel manipulation and dumbbell packing using Action Chunking Transformer.

Bibtex

@article{kawaharazuka2026mevion,
  title={{MEVION: Low-Cost Open-Source Data Collection System for Powerful and High-Speed Dual-Arm Manipulation}},
  author={Kento Kawaharazuka and Yoshiki Obinata and Hirokazu Ishida and Jihoon Oh and Temma Suzuki and Shintaro Inoue and Keita Yoneda and Ayumu Iwata and Kei Okada},
  journal={IEEE Robotics and Automation Practice},
  year={2026},
}