Rapid Assemblers for Voxel-Based VLSI Robotics
Abstract
The goal of this research is to develop scalable fabrication methods and design tools for assembling large scale robots out of billions of micro scale building blocks (voxels) arranged on a regular 3D lattice. Traditionally, robotic systems are constructed of custom-designed and fabricated components with arbitrary geometric and functional interfaces. The arbitrary nature of such integrations makes it difficult to design, simulate and fabricate complex 3D machines, as well as to repair, adapt and recycle existing machines, and to perform rigorous design automation. In contrast, electronic systems have reached very large scale integration (VLSI) levels through consistent definition of elementary building blocks, interfaces, and design rules. Biological systems, too, construct, repair, adapt and recycle large scale and complex 3D organisms by combining and recombining a relatively small repertoire of building blocks types - cell types, proteins types or amino acids types, depending on the scale of interest. In this proposal we envision a new robotics paradigm where machines are comprised of very large scale integration of billions of small scale building blocks (voxels). If each voxel has the size of about 300 microns, then the resulting assembly could look as smooth to the eye as a typical XGA LCD display - at about 85 Pixels Per Inch (PPI). The voxels will be self-aligning and interlocking, resulting in integrated and precise large scale integrations. Some voxels can be passive and very cheap, while others might be active sensors, actuators, computational and power components used more sparingly. The range of types of voxels available will define the range machines that can be fabricated.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 12, 2014
- Accession Number
- ADA624831
Entities
People
- Cheryl Perich
- Fnu Apoorva
- Hod Lipson
- Jonathan Hillar
- Rob Maccurdy
- Steve Bernard
Organizations
- Cornell University