Plant-Human Embodied Biofeedback (pheB)
A Soft Robotic Surface for Emotion Regulation in Confined Physical Space
“pheB,” is a robotic “plant-human, embodied, biofeedback” system to support the wellbeing of human inhabitants in confined, physical spaces. This surface aims to increase users’ emotion regulation and foster connections with nature by visualizing the internal states of plants through tactile, expressive movement and a computationally derived biophilic design. Unlike 2D biofeedback visualization models currently in use, our research explores mindfulness practices through immersive, tangible interactions to try and increase the therapeutic effectiveness. pheB is designed to provide therapeutic support manifested as gentle, responsive behaviors of the soft robot, shaped by computational transformations of electrical signals emitted by plant and human physiological states.
Full paper can be read here:
Our key research questions are two: (1) Do tangible, embodied interaction modalities increase the effectiveness of mindful breathing practices and biofeedback?; and (2) Do tangible, embodied representations of a plant’s internal states amplify the restorative effects of nature and foster meaningful plant-human relationships through a more expressive interaction?
Motivation
Humans are spending considerably less time outdoors. Looking to the future, society is preparing for the possibility of living in extreme environments on Earth and other planets. Such “homes” may be spatially confined (capsule habitats, like submarines or spacecrafts), or microhomes (dwellings less than 200 square feet). Evidence shows that confined spaces can stress inhabitants due to separation from nature and prolonged social isolation which negatively impacts emotional wellbeing.
Embedding ecologically-rich, interactive biofeedback interfaces into domestic architecture can possibly enrich the experience of living in a confined, isolated space.
Plant Bio-signals
We performed several experiments to see what stimuli created a shift in the electrical signal output from the plant and measured the voltage from the plant's signal shift in response to light detection, human touch, and the presence of water. For example, when the lights are turned on, the voltage output from the plant increases. These fluctuations in the signal will be incorporated in the system and moderate the behavior of the surface, for example by increasing the speed at which it inflates. This allows a human to see the internal experiences of the plant life through its electrical expressions.
Prototyping
The behavior of the robot is controlled in part by the shape and volume of the pre-determined cells, as defined by the design of the physical prototype as well as by the air that is fed into the pneumatic device. We created an algorithm using Rhinoceros 3D/Grasshopper that enables us to produce a wide range of digital models to cast different surface designs. In addition to a variety of pattern typologies, the algorithm made it so that as the prototype continues to develop, we can easily change the size and shape of the overall surface. The algorithm also allowed us to manipulate the underlying grid to produce desired inflation patterns, through areas of compression and expansion.
Biophilic Design & Airflow Diagramming
We also reinforced the biophilic nature of the surface by using images from nature as a generative tool for the design of the inflation patterns. We made bitmap images of natural features (e.g., size and placement of trees in the forest) and imported them into our CAD software. We then overlaid the images with a series of vectors. We applied an attraction force towards these curves using our Grasshopper script to generate a pattern that loosely corresponds to the identified features.
To explore how we might change the behavior of the prototype through the design of different air channel pathways, we produced a series of diagrams to evaluate strategies to inflate the device through different channel typologies. This series of diagrams helped us to consider the many possible inflation permutations using the same underlying grid.