Experiments in design synthesis when behavior is determined by shape

  • Authors:
  • Eric Schweikardt;Mark D. Gross

  • Affiliations:
  • Cornell Computational Synthesis Lab, Department of Mechanical and Aerospace Engineering, Cornell University, Ithaca, USA;Computational Design Lab, School of Architecture, Carnegie Mellon University, Pittsburgh, USA

  • Venue:
  • Personal and Ubiquitous Computing
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

As we rapidly approach the day of transitive materials, made of individual elements that sense and actuate and can be programmed and reprogrammed, it is time to think about how to design things using these new materials. Our roBlocks construction kit toy teaching children about emergent behavior in complex systems serves as an example for investigating the challenges of designing things made of transitive materials. The roBlocks kit comprises heterogeneous modular robotics components that exhibit modularity, one-to-one mapping between form and behavior, and non-hierarchical control; and these features make it appropriate for experimenting with emergent behavior. However, as the numbers of robotic components scales to the orders of magnitude needed to consider them as material these same features also make it difficult to apply traditional methods to design constructions with desired behaviors. To understand this design space we built, the Erstwhile Agent that uses an evolutionary approach to automatically synthesize roBlocks constructions to meet specified desiderata.