Programmable Robotic Self Assembly

We are studying the underlying mathematics of self assembly in terms of geometry, physics, communication and control. We can cast the problem of self assembly in terms of parallel algorithms and, for example, produce local assembly rules that produce any desired global shape. We are building a large-scale distributed robotic testbed with "programmable" parts that can embody the assembly and self-organization rules we are studying.  Funded by an NSF CAREER Award. Further description, photos, videos and papers here.

Verification of Cooperative Control Algorithms

A large-scale distributed robotic system consists of a set of heterogeneous vehicles and robots all concurrently executing a variety of tasks to acheive some global behavioral specification. We are developing design tools based on graph grammars and hybrid systems theory that allow the operator of the system to (1) specify local interactions and system goals; (2) formally verify that local interactions acheive the goals, or show why they do not; (3) automatically generate executable code and communications protocols from local interaction rules.

DNA Machines

A single strand of DNA exhibits complex behavior as it folds up upon itself due to Watson-Crick base pairing. With multiple strand types it is possible to create DNA motors, memory devices and logic gates. We are focusing on the problem of sequence synthesis: given a logical description of the desired behavior of a DNA folding reaction, automatically generate nucliotide sequences to produce the given reaction. Funded by the University of Washington's Royalty Research Fund.

Microscale Self Assembly

With Karl Böhringer We are building a 3D self assembling system using MEMs-scale tiles whose wetabilities can be controlled locally by the tiles themselves. The project involves design and fabrication of the parts, mathematical modeling and large-scale simulation of their collective behavior, and the development of distributed algorithms to control the aggregates they form. Funded by NSF grant number 0501628.

For further information, please see the publications page.