Organizations: Department of Mechano-Informatics, Graduate School of Information Science and Technology, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8656, Japan · AI Center, Graduate School of Information Science and Technology, The University of Tokyo, Japan
Reconfigurable robots offer significant potential for adapting to diverse tasks; however, conventional centralized architectures often require dedicated actuators for each module, leading to substantial increases in overall system weight, volume, and cost. To address these challenges, this paper presents the "Detachable Wire Drive," a reconfigurable robotic system that enables the sharing of heavy and expensive actuators across various morphologies. The core of this system is the "Wire Detach Unit," a mechanism designed to physically split and reconnect wire drive paths, allowing motors to be consolidated into a common base unit. We demonstrate the versatility of this approach by developing a 2-DOF rigid arm, a continuum arm, and two distinct grippers, all of which are interchangeably attached to, and driven by, a single shared actuator set. Experimental results validate the mechanical reliability of the detachment process and the control framework's ability to seamlessly manage transitions between configurations, highlighting a path toward more efficient and multi-functional robotic systems.
Figures & tables
Fig. 1: The two types of arms and grippers developed in this study, and the Wire Path Splitting Mechanism (center of the figure) that connects and disconnects them. The two arms are shown attached to the common upper motor-unit.
Fig. 2: Conceptual diagrams of conventional centralized reconfigurable robots and the proposed reconfigurable wire-driven robot system using a wire path splitting mechanism. Conventional systems require actuators for each module, making it difficult to achieve both morphological versatility and cost/weight savings. In contrast, the proposed system consolidates actuators in a common base unit and transmits power from there, enabling actuator sharing across a wide range of morphologies and achieving both goals. Additionally, using the transmission medium for joint state estimation avoids complex sensor connection reconfigurations during morphological changes.
Fig. 3: An overview of the reconfigurable wire-driven robot system developed in this study, utilizing the Wire Detach Unit. The system comprises a central motor-unit with motors and electronics, two distinct arm types (a 2-DOF rigid arm and a continuum arm), and two end-effector types (a parallel gripper and a multi-fingered gripper). These components can be interchangeably combined and are driven and controlled by shared actuators.
Specification
Value
wire tensil strength
1000 N
wire diameter
1 mm
wire type
Vectran [ 16 ]
winding pulley radius
15 mm
winding pulley width
10 mm
maximum winding length with same radius
942 mm
TABLE I: Specifications of the Wire Detach Unit.
Fig. 4: Design of the Wire Detach Unit. It consists of a winding pulley to take up the wire, a wire alignment pulley to ensure stable winding, a relay point to guide the wire, and convex-concave features for torque transmission. The convex-concave features have rounded edges to aid smooth interlocking. The entire unit is housed in a sector-shaped outer shell, allowing up to 6 units to be arranged in a circular pattern for collective attachment and detachment. This mechanism enables the seamless connection and disconnection of both the physical structure and the wire-based power transmission path.
Fig. 5: Design of the motor-unit and the motor-arm detacher. The motor-unit (left) houses the motors and features the female side of the detacher. The arm (right) is equipped with the male side, which includes multiple Wire Detach Units. The detacher employs a latch mechanism with pins on the male part and a rotating outer ring with pin holders on the female part, enabling quick and secure attachment and detachment.
Fig. 6: Design of the arm-end-effector detacher. The male part is on the arm side and the female part is on the end-effector side. A single Wire Detach Unit is integrated into each side the mechanism to transmit power to the end-effector. Similar to the motor-arm detacher, it uses a pin-and-locking-ring mechanism for quick attachment and detachment.
Fig. 7: Design of the 2-DOF rigid arm. It features a shoulder joint (joint1) and an elbow joint (joint2). Three wires for arm actuation are routed from the base and attached to the second link. A separate wire for the end-effector is guided through a Bowden cable to the detacher at the arm’s tip.
Fig. 8: Design of the continuum arm. The arm is constructed from a series of identical units, each composed of a central spring and two triangular plates. Wires for actuation run through the structure to the tip. A flexible sheet wraps around the springs to provide structural integrity and constrain excessive twisting.
Fig. 9: Design of the parallel gripper. It features two jaws that move in parallel. A single wire from the attached Wire Detach Unit drives the closing motion, while an elastic band provides the opening force. Bearings are used at sliding and bending points to ensure smooth operation.
Fig. 10: Design of the multi-fingered gripper. It has three underactuated fingers, each with three joints, driven by a single wire for closing and elastic bands for opening. This allows adaptive grasping. A common pulley inside the arm-ee detacher couples the three finger wires, requiring a slight modification to the detacher’s Wire Detach Unit.
Fig. 11: State transition diagram for the controller. The system can be in one of four states: FREE (motors off), ATTACH (active engagement assistance), CALIBRATE (joint angle estimation reset), and CONTROL (closed-loop operation). The diagram shows the permitted transitions between these states, managed according to operator commands and system status.
Fig. 12: Block diagram of the joint control system. The controller takes a reference joint angle qref and calculates motor torques Tref . The process involves a Joint Angle Follower, which computes reference wire tensions fref from the joint angle error, and a Joint Angle Estimator, which estimates the current joint angles q^ from motor encoder readings θ .
Fig. 13: Results of the strength and friction evaluation experiment for the Wire Detach Unit. Input tension was cycled from 0 to 450 N and back to 0 over 60 seconds. The deviation between input and output tension averaged approximately ± 16 N, which is sufficiently small compared to the ideal direct correspondence (red dotted line). The unit successfully withstood 450 N of tension without failure.
Fig. 14: Experiment showing the sequential attachment, control, and detachment of two different arms (rigid and continuum) from the motor-unit. The graphs on the right show the target and estimated joint angle transitions for both arms during control. In both cases, the estimated joint angles successfully track the target joint angles.
Fig. 15: Experiment showing the sequential attachment, control, and detachment of two different grippers on the rigid arm. Both grippers demonstrate the ability to properly grasp and manipulate objects according to their shape and size.
Comparison
RMSE [rad]
q vs. q^
0.087
qref vs. q^
0.145
qref vs. q
0.177
TABLE II: RMSE evaluation of the error between the target joint angle qref , estimated joint angle q^ , and actual joint angle q in the joint control state of the rigid arm. The results show that the joint angles can be estimated with sufficient accuracy from the wire length, and that there is sufficient tracking of the target joint angles.
Fig. 16: Transition of target joint angle qref , estimated joint angle q^ , and actual joint angle q in the joint control state of the rigid arm. It can be seen that q^ generally follows q , and q^ generally follows qref .
Parallel wire-driven systems are a framework in which multiple wires are connected to an end effector, and the end effector is driven by the tensions of those wires. Because they can transmit large forces, up to the breaking strength of the wires, over lengths on the order of several tens of meters, they can realize powerful actuation in large workspaces. In recent years, studies have also investigated connecting wires to external environments such as trees and columns and forming parallel wire actuation on site. However, conventional parallel wire-driven systems often assume a single end effector, and because of interference between wires and links in a shared workspace, it has been difficult to handle multiple floating links simultaneously. In this study, we realize multi-floating-link parallel wire actuation by introducing an inter-body wire. In real-robot experiments, the two bodies supported and lifted a mannequin and transferred it onto a stretcher as a rescue motion. This shows that parallel wire actuation can be extended from a single end effector to a multi-floating-link system, and indicates its potential for broader tasks such as the manipulation of large objects and rescue support.
Department of Mechano-Informatics, Graduate School of Information Science and Technology, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan
Modular small-scale robots offer the potential for on-demand assembly and disassembly, enabling task-specific adaptation in dynamic and constrained environments. However, existing modular magnetic platforms often depend on workspace collisions for reconfiguration, employ bulky three-dimensional electromagnetic systems, and lack robust single-module control, which limits their applicability in biomedical settings. In this work, we present a modular magnetic millirobotic platform comprising three cube-shaped modules with embedded permanent magnets, each designed for a distinct functional role: a free module that supports self-assembly and reconfiguration, a fixed module that enables flip-and-walk locomotion, and a gripper module for cargo manipulation. Locomotion and reconfiguration are actuated by programmable combinations of time-varying two-dimensional uniform and gradient magnetic field inputs. Experiments demonstrate closed-loop navigation using real-time vision feedback and A* path planning, establishing robust single-module control capabilities. Beyond locomotion, the system achieves self-assembly, multimodal transformations, and disassembly at low field strengths. Chain-to-gripper transformations succeeded in 90% of trials, while chain-to-square transformations were less consistent, underscoring the role of module geometry in reconfiguration reliability. These results establish a versatile modular robotic platform capable of multimodal behavior and robust control, suggesting a promising pathway toward scalable and adaptive task execution in confined environments.
Erik Garcia Oyono, Jialin Lin, Dandan Zhang
Department of Bioengineering, Imperial College London, London, United Kingdom.
Robots that can change their morphologies and behaviors for different tasks and environments hold great promise for adaptable, multifunctional systems. Modular reconfigurable robots (MRRs) can achieve such functionalities by docking and rearranging individual units, but most rely on rigid modules that lack structural compliance, resulting in limited capabilities. Continuum robots offer compliance through flexible backbones, yet they cannot self-reconfigure into task-adaptive multi-robot configurations. Here, we introduce an MRR that unifies the advantages of both architectures by combining a tensegrity-based compliant body with claw-based connection mechanisms. Each robot can manipulate and locomote independently, and multiple robots can self-reconfigure into different morphologies (e.g., chains, loops, branches) for cooperative manipulation and locomotion. We demonstrate the robots' capability across diverse tasks and environments, including coordinated object manipulation and transport, multimodal locomotion, and loco-manipulation in real-world scenarios. These results lay a foundation for adaptable and multifunctional robotic collectives, with broad potential applications in manufacturing, space exploration, and search-and-rescue operations.
Mahmud Hasan Saikot, Sydney Spiegel, Sudheera Akalanka Kariyawasam +3
Adaptive Robotics Lab, Department of Mechanical Engineering, Colorado State University, Fort Collins, CO, USA