Application of machine learning for isomorphic analysis of robotic wrist mechanisms
Jiyaul MUSTAFA, Abdullah Aamir HAYAT
Abstract. Robotic wrist mechanisms are essential for achieving dexterity, precision, and orientation control in manipulators. Traditional isomorphic analysis, based on graph-theoretic tools such as connectivity matrices, articulation points, and Wiener numbers, is effective but limited by scalability, efficiency, and the need for manual interpretation. This paper proposes a machine learning (ML) framework that represents wrist topologies as graphs and applies Graph Neural Networks (GNNs) to identify structural equivalence. A benchmark dataset was created from Tsai’s atlas of spherical and non-spherical wrist mechanisms, augmented with synthetically generated topologies. The preliminary simulation results demonstrate that the GNN-based model achieves over 95% accuracy in classifying isomorphic and non-isomorphic designs, while reducing classification time by approximately 60% compared to the Wiener number method. This preliminary study relies heavily on a data set, which may misclassify symmetrical structures. With this, the future work aims to expand the data sets, conduct detailed computational comparisons, and utilise other learning approaches.
Keywords
Robotic Wrist Mechanisms, Isomorphism, Graph Theory, Machine Learning, Graph Neural Networks, Mechanism Synthesis, Structural Analysis
Published online 6/20/2026, 10 pages
Copyright © 2026 by the author(s)
Published under license by Materials Research Forum LLC., Millersville PA, USA
Citation: Jiyaul MUSTAFA, Abdullah Aamir HAYAT, Application of machine learning for isomorphic analysis of robotic wrist mechanisms, Materials Research Proceedings, Vol. 67, pp 465-474, 2026
DOI: https://doi.org/10.21741/9781644904176-62
The article was published as article 62 of the book Climate Action and Sustainability
Content from this work may be used under the terms of the Creative Commons Attribution 3.0 license. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
References
[1] Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., Yu, P.S., A comprehensive survey on graph neural networks, IEEE Trans. Neural Netw. Learn. Syst. 32 (1) (2021) 4–24. https://doi.org/10.1109/TNNLS.2020.2978386
[2] Jalayer, R., Jalayer, M., Orsenigo, C., Tomizuka, M., A review on deep learning for vision-based hand detection, hand segmentation and hand gesture recognition in human–robot interaction, Robot. Comput.-Integr. Manuf. 97 (2026) 103110. https://doi.org/10.1016/j.rcim.2025.103110
[3] Zhou, J., Cui, G., Zhang, Z., Yang, C., Liu, Z., Wang, L., Li, C., Sun, M., Graph neural networks: A review of methods and applications, AI Open 1 (2020) 57–81. https://doi.org/10.1016/j.aiopen.2021.01.001
[4] Mustafa, J., Hasan, A., Khan, R.A., Identification of isomorphism in epicyclic gear trains – a Bocher’s approach, Mater. Today: Proc. 18 (2019) 2474–2481. https://doi.org/10.1016/j.matpr.2019.12.047
[5] Kipf, T.N., Welling, M., Semi-supervised classification with graph convolutional networks, in: Proc. Int. Conf. Learn. Represent., 2017. https://doi.org/10.48550/arXiv.1609.02907
[6] W. Yang, and H. Ding, The complete set of one-degree-of-freedom planetary gear trains with up to nine links. ASME J. Mech. Des. 141(4), (2019) 043301. https://doi.org/10.1115/1.4041482
[7] Rai, R.K. and Punjabi, S., A new algorithm of links labelling for the isomorphism detection of various kinematic chains using binary code. Mech. Mach. Theory 131, (2019) 1-32. https://doi.org/10.1016/j.mechmachtheory.2018.09.010
[8] Tsai, L.W., Mechanism Design: Enumeration of kinematic structures according to function, CRC Press, Boca Raton, Florida, USA, (2001). https://doi.org/10.1115/1.1334346
[9] Tsai, L.W., Robot Analysis: The Mechanics of Serial and Parallel Manipulators, Wiley, New York, USA, 1999.
[10] Tsai, L.W., Systematic enumeration of parallel manipulators, in: C.R. Boër, L. Molinari-Tosatti, K.S. Smith (Eds.), Parallel Kinematic Machines: Theoretical Aspects and Industrial Requirements, Springer-Verlag, London, UK, 1999, pp. 33–49. https://doi.org/10.1007/978-1-4471-0885-6_3
[11] Tahmasebi, F., Tsai, L.-W., On the stiffness of a novel six-DOF parallel mini-manipulator, J. Robot. Syst. 12 (12) (1995) 845–856. https://doi.org/10.1002/rob.4620121208
[12] Wohlhart, K., Displacement analysis of the general spherical Stewart platform, Mech. Mach. Theory 29 (4) (1994) 581–589. https://doi.org/10.1016/0094-114X(94)90096-5
[13] Meng, Z., Cao, W.-a., Ding, H., Chen, Z., A new six degree-of-freedom parallel robot with three limbs for high-speed operations, Mech. Mach. Theory 173 (2022) 104875. https://doi.org/10.1016/j.mechmachtheory.2022.104875
[14] Innocenti, C., Parenti-Castelli, V., Echelon form solution of direct kinematics for the general fully parallel spherical wrist, Mech. Mach. Theory 28 (4) (1993) 553–561. https://doi.org/10.1016/0094-114X(93)90035-T
[15] Gosselin, C., Angeles, J., The optimum kinematic design of a spherical three-degree-of-freedom parallel manipulator, ASME J. Mech. Transm. Autom. Des. 111 (1989) 202–207. https://doi.org/10.1115/1.3258984
[16] Lin, C.C., Tsai, L.W., The development of an atlas of bevel-gear type spherical wrist mechanisms, in: Proc. 1st Nat. Conf. Applied Mechanisms and Robotics, 1989, Paper No. 89-AMR-2A-3. http://hdl.handle.net/1903/4902
[17] Tsai, L.W., The kinematics of spatial robotic bevel-gear trains, IEEE J. Robot. Autom. 4 (2) (1988) 150–156. DOI: 10.1109/56.2078
[18] Trevelyan, J.P., Kovesi, P.D., Ong, M., Elford, D., ET: A wrist mechanism without singular positions, Int. J. Robot. Res. 4 (4) (1986) 71–85. https://doi.org/10.1177/027836498600400406
[19] Rao, A.C., A genetic algorithm for epicyclic gear trains, Mech. Mach. Theory 38 (2) (2003) 135–147. https://doi.org/10.1016/S0094-114X(02)00094-0
[20] Hunt, K.H., Structural kinematics of in-parallel-actuated robot-arms, ASME J. Mech. Transm. Autom. Des. 105 (1983) 705–712. https://doi.org/10.1115/1.3258540
[21] Hsu, C.H., Wu, Y.C., Automatic detection of embedded structure in planetary gear trains, ASME J. Mech. Des. 119 (2) (1997) 315–318. https://doi.org/10.1115/1.2826253
[22] Uicker, J.J., Raicu, A., Identification of equivalence of kinematic chains, Mech. Mach. Theory 12 (1975) 375–383. https://doi.org/10.1016/0094-114X(75)90037-3
[23] Chen, D., Shiue, S., Topological synthesis of geared robotic mechanisms, ASME J. Mech. Des. 120 (2) (1998) 230–239. https://doi.org/10.1115/1.2826963.
[24] Romiti, A., Raparelli, T., Sorli, M., Robot wrist configurations, mechanisms and kinematics, in: Robotics in Alpe-Adria Region, Springer, Vienna, Austria, 1994, pp. 44–48. https://doi.org/10.1007/978-3-7091-9346-4_8
[25] Pan, W., Li, R., The equivalent mechanical model of topological graphs and the isomorphism identification of kinematic chains, J. Mech. Robot. 16 (7) (2024) 071014. https://doi.org/10.1115/1.4063869
[26] Hou, W., Li, Y., Wang, C., et al., FrameGraph: A scalable performance evaluation method for frame structure designs using graph neural network, J. Mech. Des. 146 (12) (2024) 121703. https://doi.org/10.1115/1.4065612
[27] Esmail, E.L., A universal kinematic analysis of geared mechanisms. Journal of the Brazilian Society of Mechanical Sciences and Engineering 39, (2017) 2253–2258. http://dx.doi.org/10.1007/s40430-017-0711-2
[28] Mustafa, J., Hasan, A., Some application of graph theory to isomorphic analysis of epicyclic geared mechanisms, J. Inst. Eng. (India): Ser. C 102 (2021) 1149–1157. http://dx.doi.org/10.1007/s40032-021-00710-x
[29] Diao, X., Chi, W., Wang, J., Graph neural network based method for robot path planning, Bioinspir. Robot. 4 (2024) 100147. https://doi.org/10.1016/j.birob.2024.100147
[30] Almeida, J.D., Schydlo, P., Dehban, A., Santos-Victor, J., Sensorimotor graph: Action-conditioned graph neural network for learning robotic soft hand dynamics, arXiv preprint arXiv:2107.08492, 2021. http://dx.doi.org/10.1109/IROS51168.2021.9636377
[31] Wang, M., Wang, J., Zhang, J., et al., Physics-informed graph learning for shape prediction in robot manipulation of deformable linear objects, Chin. J. Mech. Eng. 38 (55) (2025) 1–13. http://dx.doi.org/10.1186/s10033-025-01299-3
[32] Wei, R., Liu, Y., Dong, H., et al., A graph-based hybrid reconfiguration deformation planning for modular robots, Sensors 23 (18) (2023) 7892. https://doi.org/10.3390/s23187892
[33] Sharma, S., Purwar, A., A machine learning approach to solve the Alt-Burmester problem for synthesis of defect-free spatial mechanisms, J. Comput. Inf. Sci. Eng. 22 (2) (2022) 021003. https://doi.org/10.1115/1.4051913
[34] Purwar, A., Chakraborty, N., Deep learning-driven design of robot mechanisms, J. Comput. Inf. Sci. Eng. 23 (6) (2023) 060811. https://doi.org/10.1115/1.4062542
[35] Hayat, A.A., Megalingam, R.K., Kumar, D.V., Rudravaram, G., and Nansai, S., Sensitivity analysis, synthesis and gait classification of reconfigurable Klann legged mechanism, Mathematics, vol. 12, no. 3, p. 431, 2024. https://doi.org/10.3390/math12030431
