Background: Recent progress in molecular and developmental biology have revealed the ubiquity of self-assembly as a fundamental mechanism underlying processes that organize and maintain the proper functioning of living systems at all length scales. Self-assembling systems admit a large, possibly infinite, variety of global structures through the interaction of a well-defined number of components. The expected average properties of the emerging structures are not only determined by the way the components are linked to each other but may also depend critically on the dynamics of the self-assembly process. This means that the structural properties of a large relevant class of self-assembling systems can only be understood by combining theoretical analysis and experiments that take the physical embodiment into account. This proposal rests on three basic assumptions: (1) self-assembly processes at all scales are subject to universal organizational principles; (2) insights into the logic underlying such principles can be gained by carefully studying the information flow networks defined by the parts composing a self-assembly system, and between the system and the environment in which it is embedded; and (3) morphology of the parts plays a crucial role in any self-assembly process – whereby morphology represents the combination of geometrical parameters (size, angles, symmetries, shape, etc.), physical properties (mass, moment of inertia, charge, magnetic dipole, etc.), and functional characteristics (type of receptors, location of receptors, and effectors).

Project summary: This project aims at systematically studying how the structures emerging from a macroscopic (mm and cm scale) 2D self-assembly process depend on the behavior and morphology of the interacting parts and their coupling with the environment. We will pursue this goal by combining modeling techniques (based on state-of-the-art physics-based simulators), experimental studies (employing a 3D printer for rapid proto-typing), and descriptive tools for the statistical and information-theoretic analysis of SA systems. The accomplishment of this goal will convey further benefits:

1. A procedure to effectively design – given a desired task-environment – the morphology and behavior of the parts of a macroscopic 2D self-assembly system.

2. A quantitative understanding of the relation between the morphology and local behavior of the parts, the global behavior of the assembled structure, and the task-environment of a given system.

3. A general and scalable model of the relationship between morphology and yield ratio (i.e. the probability of a system assembling a particular configuration) that can be applied to any 2D self-assembly system – including microscopic systems.

Importance: Living systems are the best proof of the power of self-assembly. While biological inspiration is now widely used in the development of technology, in particular robots – an example is neurally inspired robotics – engineers seem to shy away from exploiting processes of self-organization, emergence, and self-assembly. We strongly suspect that one of the reasons is a lack of understanding and in particular the fact that effective design principles for self-assembly that would enable scientists to “design for emergence”, are almost entirely missing. We are confident that with this project, we will be able to deepen our understanding of self-assembly and develop design principles that may eventually empower technology developers to build applications whose adaptive capabilities are far beyond current systems.

Keywords: Self-assembly systems, self-assembly simulations, experimental model validation, morphology, information theory, modular robotics, embodiment.