Programmable Assembly With Universally Foldable Strings (Moteins)

Abstract

Abstract-Understanding how linear strings fold into 2-D and 3-D shapes has been a long sought goal in many fields of both academia and industry. This paper presents a technique to design self-assembling and self-reconfigurable systems that are composed of strings of very simple robotic modules. We show that physical strings that are composed of a small set of discrete polygonal or polyhedral modules can be used to programmatically generate any continuous area or volumetric shape. These modules can have one or two degrees of freedom (DOFs) and simple actuators with only two or three states. We describe a subdivision algorithm to produce universal polygonal and polyhedral string folding schemas, and we prove the existence of a continuous motion to reach any such folding. This technique is validated with dynamics simulations as well as experiments with chains of modules that pack on a regular cubic lattice. We call robotic programmable universally foldable strings "moteins" as motorized proteins.

Programmable Self-assembly with Chained Soft Cells: An Algorithm to Fold into 2-D Shapes

“…In order to illustrate the advantages of our system composed of cells with non-uniform softness settings, we also run the algorithm with uniform softness settings (which is the case in state-of-the-art systems [6][7][8][9][10]). This result is also plotted in Figure 4(b).…”

“…The algorithm presented here is well-suited to be combined with a space- filling algorithm such as the one presented in [7]. Then, a robot's morphology may be automatically constructed by approximating its surface using the method described here and by specifying the required filling of the body (for instance according to functional constraints such as weight, etc.).…”

“…Through the use of four tile-types with different magnetic patterns, their system could approximate any two dimensional shape. More recently, this work was extended to three dimensions, along with algorithms to approximate any three dimensional shape [7]. Inspired by this work, Knaien et al [8] developed a system called "MilliMoteins" that consisted of a chain of electropermanent magnetic motors that could dynamically fold into different configurations.…”

“…In order to approximate any shape with high accuracy, we put emphasis on matching a 2D hamiltonian path of target shapes. As the search for a hamiltonian path through shapes is part of other research [7], we avoid this step and focus on surface shape matching here. Hence, we provide an algorithm that can approximate the surface of any 2-D shape with high accuracy using a programmatically self-assembling system composed of soft modules.…”

A Stochastic Approach to Shortcut Bridging in Programmable Matter

“…In passive systems, computational units cannot control their movement and have (at most) very limited computational abilities, relying instead on their physical structure and interactions with the environment to achieve locomotion (e.g., [28,1,21]). A large body of research in molecular self-assembly falls under this category, which has mainly focused on shape formation (e.g., [13,8,27]). Rather than focusing on constructing a specific fixed target shape, our work examines building dynamic bridges whose exact shape is not predetermined.…”