horizontal line
Living Machines Home navigation buttonseparator lineOverview navigation barseparator linePeople navigation buttonseparator lineProjects navigation buttonseparator linePublications navigation bar

Developmental Nervous Systems for Tensegrity Robots
a prototype tetrahedron tensegrity structure
People: Aaron Edsinger, Miguel Ferreira (UROP), Naveen Goela (UROP)


Our research is concerned with the autonomous development of physically embodied robots. We are looking at simple invertebrates such as the leech and the jellyfish to provide models of neural and morphological development. By taking a computational approach to such models, we hope to confirm experimental biological findings and to elucidate the underlying set of tools that nature employs in the development of an organism. Currently, mechanisms of neurogenesis and morphogenesis, along with neural plasticity, are the primary areas of investigation. In addition, we are looking at homeostatic systems and artificial chemistries as components of the developmental program. These models of living systems can provide a framework within which tools such as evolutionary search can find system architectures and morphologies for the robots.

Our hope is that by situating the robots, equipped with real-time sensors and effectors, in a complex environment, the richness of the world can guide the artificial organisms development towards interesting behaviors. The robots being built for this purpose are based on the structural phenomena of tensegrity. Tensegrity structures, discovered in part by Buckminster Fuller, contain isolated compression beams seemingly floating in a sea of elastic tension provided by cables. This highly efficient building technique is widespread in biological systems, from the cell membrane to the musculature system. The tensegrity robots under construction are to be highly compliant and driven by force-actuation. As such, they can provide rough models for organisms such as the leech and jellyfish.