CAREER: Programmed Robotic Self Assembly

University Of Washington, Seattle WA

Investigators

Abstract

Self-assembly (SA) occurs when many building blocks are placed in an environment that thermodynamically favors their forming structured aggregates. For example, capillary forces between millimeter scale tiles floating on a liquid-liquid interface produce regular arrays of various shapes as the system is gently shaken. Similar phenomena play crucial roles in the assembly of viruses, organism development, growth of silicon crystals, electronics, pathogen detectors, artificial tissue, ad hoc wireless networks and reconfigurable robotic devices. However, most examples of engineered SA usually produce rather simple, regular arrays of parts. Engineering a system that self assembles something as complicated and irregular as, for example, a virus is well beyond our capabilities both theoretically and technologically. The PI proposes to address this problem by supplying the parts to be assembled with active control over the binding interactions in which they participate, developing a theory and practice of Programmed Robotic Self-Assembly. In particular, the PI proposes to implement programmed robotic SA in a test-bed where programmable parts have micro-controllers, local sensors and mechanical or magnetic latches. The idea is to approximate organic SA where changes in molecular conformation (shape) guide the assembly process. A task that the test-bed should be able to perform is: "Given a desired assembly specification input by the user, download programs onto each robotic part so that, when they are placed in a suitable environment, the parts self assemble into copies of the desired assembly." The success of this agenda will yield a useful class of robotic self-assembling structures as well as a theory for their design and operation. The proposed research involves theoretical models from computer science, dynamical systems and control, robotics and embedded systems. Thus, the interdisciplinary education of students of all levels is required. To address this need, the PI will develop two new courses, on Self Organizing Systems and Specification and Control, and interdisciplinary collaboration aimed at creating graduate and undergraduate students well versed in computer science, robotics, controls and the information-rich physical settings in which these fields are increasingly applied. The PI recognizes that not all students have equal experiences leading up to and while in college. Therefore, the PI will collaborate with the University of Washington College of Engineering's Mathematics Engineering Science Achievement program to employ high school student interns in the PI's laboratory. Furthermore, to make the results of the proposed research and curriculum development accessible to the public, the PI proposes to create interactive demonstrations of SA and robotics more generally. This material will be made available on the web as well as in the PI's lab and will be used by the PI and his students for education and outreach in conjunction with the University of Washington Electrical Engineering Department's K-12 outreach program