A machine learning-aided approach to the rapid design of kirigami-inspired soft deployable structures

Shape-morphing soft structures that spontaneously transition from planar to 3D shapes are transformative technologies with broad applications in soft robotics and deployable systems. However, the high dimensionality causes designing soft deployable structures challenging, which used to be a process of trial and error with complex local actuation and fabrication. We report a rapid design approach for fully soft structures that can achieve targeted 3D shapes through a fabrication process that happens entirely on a 2D plane. To relax the need for local actuation, we develop a much-simplified planar fabrication approach that combines the strain mismatch in the composite structure and kirigami designs. To expedite the design process and explore the capability of such a much-simplified fabrication approach, we develop and apply a symmetry-constrained active learning approach to optimize the design parameters so that target 3D shapes can be achieved. By exploring the nonlinear interplay between kirigami patterns and strain-mismatch, we can create a wide range of 3D shapes. We demonstrate the effectiveness of the rapid design procedure via a range of target shapes with increasing shape complexity, such as those inspired by the peanuts and pringles to flowers and pyramids. Tabletop experiments are conducted to fabricate the target shapes. Tabletop-controlled experiments and finite element simulations agree in these design examples.