A Walk in Nature

Exploring the Creative Potentials of a Generative Design Tool for Soft Robotic Surfaces that Foster a Connection with Nature.

This work explores how biophilic design (based on the theory that humans have an evolutionary desire to connect with nature) can be applied to interactive systems. We describe a generative design tool that produces a bio-inspired digital prototype of a soft robotic surface. In our workflow, participants walked outdoors to photograph natural features in their environment. The photographs are used to generate a set of corresponding designs. As early validation, we conducted a small user study to test the robustness of our tool with different photographic inputs. We also collected feedback on how the workflow influenced participants’ perceptions of the design outcomes. The results guided the fabrication of a new physical prototype and demonstrated the utility of our approach, which aims to inform the design of future biophilic cyber-physical systems.

Full paper can be read here:

Sabinson, E., & Green, K. E. (2023). A Walk in Nature: Exploring the Creative Potentials of a Generative Design Tool for Soft Robotic Surfaces that Foster a Connection with Nature. Proceedings of the 15th Conference on Creativity and Cognition, 185–199. https://doi.org/10.1145/3591196.3593367

This work explores how soft robotic systems, which have an organic quality, can be used to evoke experiences that remind us of nature. Previous user studies on the system have found that soft robots can cause experiences of discomfort or an ‘uncanny valley’ like effect, due to the synthetic materiality and movement of the surfaces, which appears to be lifelike, but not quite natural. To address these reactions, we explore how participation in the creative process through a computer-aided design (CAD) workflow might promote positive feelings and a sense of attachment to soft robotic surfaces and ameliorate negative perceptions.

Current rectangular prototype of A soft robotic surface panel

We explore how we might increase positive perceptions of the system through designs that exemplify principles of biophilic design, a strategy informed by the theory of Biophilia to provide connection to the natural within the built environment. This is pertinent, as research suggests soft robots that are perceived as more natural are preferred.

Our generative design tool enables people without CAD skills to guide the design of a bio-inspired soft robotic surface. In the proposed workflow, people take photographs of nature to identify interesting shapes. The images are used to automatically generate a set of personalized designs for a “digital prototype” of a soft robotic surface.

Maps created using GIS from each participant’s walk in nature, documenting the interesting shapes they observed and photographed.

An overview of the Generative Design Tool created to translate photographs into designs for soft robotic surfaces using Grasshopper.

Image processing results, using three filters (contrast, extract hue, and edge detection) to generate the edge for the boundary surface.

A set of designs was generated for each photograph using a genetic algorithm, which employed k-means clustering to create a representative design.

This process aimed to provide participants with a sense of 'genetic design history' for artificially produced products and to enhance their biophilic sensibility. Three generations of designs were produced by Biomorpher interactive genetic algorithm with selections highlighted to show “parents” used in the design process from each generation.

PARTICIPANT FEEDBACK

FABRICATION OF A BIOPHILIC SOFT ROBOTIC SURFACE USING THE GENERATIVE DESIGN TOOL

The Generative Design Tool was also used to create Computer-Aided Machining files for 3-D printing or CNC-milling the parts to create the molds used to fabricate a soft robotic surface.