Architectural models, on the other hand, are moving towards ever-increasing levels of resolution. BIM is resulting in the digital representation of almost every single building component, down to the hardware (and even at early stages of design).

Thus, I propose adapting shape displays to a new end – one that mediates between virtual display and physical interaction,  BIM model and architectural designer, two-dimensional representations and three. Shape displays will take the form of a simple, primitive shape. This shape can be molded, stretched, skewed, etc. These basic transformations will then be correlated with a connected digital model. The user will isolate a portion of the model to manipulate (or choose the entire model), and then use the shape display to perform simple transformations. In this way the user can interact directly (i.e., tangibly) with the digital model, whether he or she is manipulating one-, two-, or three-dimensional constraints. Once the transformation is complete, the shape display re-sets back to its primitive form (or a different one – whichever the user chooses).

In the end, the digital model is allowed to be as detailed as it needs to be, unbound by the resolution restriction of the interface. The interface makes possible a tangible manifestation of the simple modeling transformations that designers currently perform through clicks and menu options. The interface and the model can connect, but are not bound to one another.

Alternatively, once the discrete model element is chosen, the parameter clay could project a simplified version of this form, upon which the user would then act.

Our ancestors used the bottoms of their feet to tell them about the world – as our only constant point of contact with the earth, our feet inform us about terrain, warn us about potential danger, and keep us balanced and upright. We see evidence of this in the fact that the bottom of the foot has one of the most dense concentrations of neural receptors in the body. These days, however, we miss that fact: Our feet are all too often cushioned and insulated from any amount of sensory feedback whatsoever.

PneuiFeet attempts to address this by introducing pneumatically-actuated sensations directly beneath the foot. In our early iterations of the project, we introduced this idea through the use of a carpet – a bathroom or kitchen rug, for instance, that would relay information to you in the morning while you got ready for the day. We later made the rug mobile by converting it to a pair of slippers: Now the capability travels with you. Finally, we settled on a silicone insole, tying into an existing market and permitting the technology to traverse outside the home.

Design + Build

Building on the work of the Dynamic Texture Change project (Yao et al), we began by experimenting with different chamber formations: point, line, arc, spiral, etc. We tested the resulting actuations with both our feet and our hands, finally settling on a chain of metal beads that was easy to prototype, simple to remove from the solidified silicone, and effective at creating a sensation (even under pressure).

Later, we moved from rectangular casts to an actual slipper shape. We created molds for a women’s size 6 and a men’s size 10.5 slipper, although ultimately we only cast iterations of the former. The molds were waterjet cut from 1/4″ thick polycarbonate and glued together to create the desired thickness. Two different configurations were cast: One with three long chambers running parallel with the feet, and one with six shorter chambers running perpendicular to the feet.

Finally, the insole was inserted into a modified shoe for testing. Holes were drilled in the shoe to allow the entrance of pneumatic tubes from the pump and microcontroller.

For visual purposes, we have demonstrated the actuation of the insole outside of the shoe.

Applications

The applications for this technology are manifold. From AR to VR, to medicine, fitness, and social media, haptic feedback is increasingly being realized as critical to our digital interactions. Below we present three possible scenarios:

1. Reflexology

The study of reflexology aims to provide relaxation and wellness through foot massage, based on the theory that there are reflex points on the feet mapped to every point in the body. This is an ancient practice that enjoys modern relevance in everything from foot massages to acupuncture. If the chambers of PneuiFeet can be made small enough (perhaps through the aid of a rigid “focusing” layer that would not allow bubble expansion), then these insoles could be programmed to target different points in the body, as needed.

2. Diabetic sensory neuropathy

This gradual nerve dysfunction is a common side effect of diabetes. It results in loss of sensation in the sensory extremities of the body, including the feet. While treatment for diabetes focuses largely on glucose levels, some patients have reported symptom alleviation through the use of biofeedback (“Diabetic Neuropathy”), such as that provided by PneuiFeet. In addition, patients could use PneuiFeet to monitor the neuropathy, reporting at regular intervals whether the sensations are increased or decreased (without having to see a specialist).

3. Navigation

Finally, we see great potential for the use of PneuiFeet in non-visual sensory navigation. Research has shown that even simple stimulation of certain places on the bottom of the feet can have an impact on “sensory steering” (Zehr et al). Linked with GPS and a predetermined route, PneuiFeet could provide turning information by actuating in individual shoes. PneuiFeet could also increase or decrease the frequency of its inflation in order to encourage the user to speed up or slow down, perhaps based on traffic conditions or weather reports. In addition, PneuiFeet could connect with a variety of sensors or public infrastructural systems to offer information for the visually impaired, such as when a crosswalk signal is activated.

References

“Diabetic Neuropathy (Nerve Damage) – An Update.” Diabetic Neuropathy. N.p., n.d. Web. 26 Oct. 2015.

Yao, Lining, Ryuma Niiyama, Jifei Ou, Sean Follmer, Clark Della Silva, and Hiroshi Ishii. “PneUI: Pneumatically Actuated Soft Composite Materials for Shape Changing Interfaces,” 13–22. ACM Press, 2013. doi:10.1145/2501988.2502037.

Zehr et al.: Cutaneous stimulation of discrete regions of the sole during locomotion produces “sensory steering” of the foot. BMC Sports Science, Medicine, and Rehabilitation 2014 6:33.

Thus, I propose Sketchbox: the expansion of current sketch surfaces into pneumatically-controlled, interactive, and malleable 3D objects. These surfaces will be directly connected to digital models. They can represent actual objects in the model, parameters of individual objects, or relationships between objects. In this way the designer is permitted to interact spatially with the model, even if the manipulation itself isn’t necessarily spatial.

For instance, imagine the designer is modeling a building with extensive louvers on the facade. He wants to test out various sizes of louvers and their aesthetic impact on the building. He sketches a circle onto the sketchpad; the circle extrudes into a cylinder; the designer then uses this object as a knob to increase or decrease the size of the louvers. This is a tactile interaction with one-dimensional data.

In another scenario, if the designer wished to study different proportions of a certain design, he or she could sketch a rough square onto the Sketchbox surface. This sketch would extend upward into a box. The designer could then stretch the resultant object along one or multiple axes, with the digital model following suit. The object itself needs to have no resemblance to the digital design; what is being translated is simply a transformational command (tangibly indicated).

Finally, in the most direct interaction method, the designer may sketch an actual object onto the surface of the Sketchbox. This sketch could then be extruded into three dimensions, with a corresponding digital object manifesting on screen.

In this way, Sketchbox provides a new form of interaction for the designer. Rather than prescribing an overly-constraining set of end possibilities for the designer to achieve, Sketchbox represents a METHOD of working, and the tools necessary to implement that method. Architectural designers are spatial by nature; their tools should encourage this.

My return to academia is an effort to pursue research I began in undergrad into human-computer interaction. I am interested in the interface between creative, subjective designers and their more rigid, deterministic computer counterparts. I am excited to study this at MIT, where I can explore this topic alongside cutting-edge work being done in both computer science and cognitive science.

I have extensive experience with Rhino, Grasshopper, and associated plug-ins for Grasshopper, as well as other CAD software and production programs (such as Adobe). I have some experience with Arduino, Processing, and Python. On the fabrication side, I’m fairly well versed in laser cutting and basic woodworking, and I have some experience in casting (concrete and plaster) and mold-making. Pretty terrible with a hammer, though.