Key points are not available for this paper at this time.
While the intensity, complexity, and specificity of robotic exercise may be supported by patient-tailored three-dimensional (3D)-printed solutions, their performance can still be compromised by non-optimal combinations of technological parameters and material features. The main focus of this paper was the computational optimization of the 3D-printing process in terms of features and material selection in order to achieve the maximum tensile force of a hand exoskeleton component, based on artificial neural network (ANN) optimization supported by genetic algorithms (GA). The creation and 3D-printing of the selected component was achieved using Cura 0.1.5 software and 3D-printed using fused filament fabrication (FFF) technology. To optimize the material and process parameters we compared ten selected parameters of the two distinct printing materials (polylactic acid (PLA), PLA+) using ANN supported by GA built and trained in the MATLAB environment. To determine the maximum tensile force of the exoskeleton, samples were tested using an INSTRON 5966 universal testing machine. While the balance between the technical requirements and user safety constraints requires further analysis, the PLA-based 3D-printing parameters have been optimized. Additive manufacturing may support the successful printing of usable/functional exoskeleton components. The network indicated which material should be selected: Namely PLA+. AI-based optimization may play a key role in increasing the performance and safety of the final product and supporting constraint satisfaction in patient-tailored solutions.
Rojek et al. (Sun,) studied this question.