Latest Papers

ASME Journal of Mechanisms and Robotics

  • Discrete Cosserat Method for Soft Manipulators Workspace Estimation: An Optimization-Based Approach
    by Walid A, Zheng G, Kruszewski A, et al. on August 11, 2021 at 12:00 am

    AbstractThis article investigates the workspace estimation of soft manipulators. Given a configuration of such a soft robot, with the bounded actuators, the discrete Cosserat method is adopted to deduce the mathematical model of soft manipulators, based on which an optimization-based approach is proposed to estimate the workspace. Implemented to various soft manipulators’ configurations, numerical simulations are provided to highlight the feasibility of the proposed methodology.

  • Design Control and Performance of a Cable-Driving Module With External Encoder and Force Sensor for Cable-Driven Parallel Robots
    by An H, Zhang Y, Yuan H, et al. on August 11, 2021 at 12:00 am

    AbstractCable-driven parallel robots (CDPRs) have the characteristic of easy deployment, which endows CDPRs with flexible workspace, freely configurable degrees-of-freedom (DOFs), and various configurations, greatly expanding their range of applications. Modular design provides excellent convenience and feasibility for deployment, which is a crucial issue of CDPR design. A highly integrated cable-driving module is designed in this paper, which includes the winding bobbin, servo motor, force sensor, external encoder, electromagnetic brake, as well as other devices. Experiments show that the maximum cable length control error is less than 0.16%, and the maximum cable tension control error is less than 8% in the back-and-forward rotation test. Furthermore, a CDPR with eight cables and six DOFs is constructed rapidly using the proposed module, whose dimension is 850 × 850 × 650 mm3. Results show that the robot’s trajectory errors are all less than 4.5 mm, and the root-mean-square-error (RMSE) is 2.1 mm. Besides, the compliance control experiments show that the robot’s tracking error in an impedance control mode is less than 2 mm, and the RMSE is 0.95 mm. Moreover, the dragging force in a teaching mode is less than 2.5 N. The proposed integrated cable-driving module could be helpful for the modular design and deployment of CDPRs.

  • A Modular Method for Mechanical Error Analysis of Planar Linkages Composed of Class II Assur Group Kinematic Chains
    by Hsu K, Chung J. on August 11, 2021 at 12:00 am

    AbstractThis paper presents a modular method for the mechanical error analysis of complex planar linkages. The topology of the linkage under investigation is decomposed into several class II Assur group kinematic chains (AGKCs) combined in a given sequence. Therefore, the mechanical error of the whole linkage can be analyzed by investigating the error propagations of adopted AGKCs in successive order. Because class II AGKCs are first served as modules, the mechanical error equations of these AGKCs in terms of each error in link lengths and joint variables can be pre-formulated and embedded in form of subroutines in any programmable language. Once the AGKCs constituting the linkage topology are identified, the corresponding subroutines are introduced to compute the error propagations in the linkage. Therefore, the presented modular approach can facilitate the analysis by concentrating on the topology decomposition instead of the algebraic derivation. Numerical examples are provided to illustrate the advantage and flexibility of the modular approach.

Development of a Hydraulic Driven Bionic Soft Gecko Toe


Geckos can climb freely on various types of surfaces using their flexible and adhesive toes. Gecko-inspired robots are capable of climbing on different surface conditions and have shown many important applications. Nonetheless, due to poor flexibility of toes, the movements of gecko-inspired robots are restricted to flat surfaces. To improve the flexibility, by utilizing design technique of soft actuator and incorporating the characteristics of a real gecko’s toe, the design of new bionic soft toes is proposed. The abilities of this bionic toe are verified using modeling and two soft toes are manufactured. One is Type A toe having varied semi-circle cross sections as the feature of real gecko toe and the other is Type B toe with a constant semi-circle cross section. The bending behaviors of the bionic toes subjected to a range of hydraulic pressure are also experimentally studied. It demonstrated that both toes can perform similarly large bending angles for the adduction (attachment) and abduction (detachment) motions. In comparisons, Type B toe exhibits larger output force, which is ascribed to the fact that at proximal section of Type B corresponds to larger volume for bearing fluid. Both toes can not only provide sufficient adhesion but can be quickly detached with low peeling forces. Finally, different curved surfaces are used to further justify the applicability of these bionic toes. In particular, the flexible toes developed also have the advantages of low cost, lightweight, and simple control, which is desirable for wall-climbing robots.
Read More
Journal of Mechanisms and Robotics Open Issues