The Material Robotics workshop of the RSS conference at MIT was an intense day. We were energized by rapid-fire 3 minute presentations, mostly on soft and squishy robots, with breaks for posters. Here is our abstract and below is the poster from our lab, all about embedding functional fibers into materials that can sense and actuate. This work features Brian Wagner’s embroidery alignment plugin, which lets you write aligned PES embroidery files (for Brother machines) in the free vector drawing software Inkscape. With this environment, it is possible to draw shapes for laser cutting alongside the fiber paths that will be in the final structure.
Yes, the actual poster about fibers is made of fibers from Spoonflower on their Performance Knit fabric.
This video shows a stretchable optical fiber that can report its own length. We have been developing sensor fibers that work with stretchable materials like athletic tape. The screen in the background shows the light intensity decreasing as the tape is stretched, and climbing back up as the tape relaxes to its original length. The intensity drop is caused mostly by the longer path length in a stretched piece of absorbing material, as shown in the animation here.
In applications where we sew the fibers onto another material, effects on the light intensity come from both the length change and from threads pushing into the fiber.
Some practical features of optical sensing compared to resistive electronic sensing is that all-polymer optical materials don’t corrode in a wet environment, cost less than metals, and don’t need a continuous conductive path. These properties will be handy when putting sensors on skin, creating disposable devices and making wearable sensors that might need to be washed.
Last week I was at the 2017 ACM CHI conference in Denver. The conference covered human-computer interaction topics ranging from accessibility, to education, to new hardware and materials for interacting with computers. When it comes to analyzing what people are doing with input devices, machine learning turned up more often than not. People presented all kinds of outputs including computer-printed food, timed release of perfumes, focused sound waves, and wearables that produce tiny electric currents to actually move your muscles. Interesting news if you’re growing tired of tapping on a keyboard and staring at a screen. It’s time to ponder how we will connect computers to soft, wearable circuits that are not guaranteed to maintain a constant shape, and come up with techniques that let people join and repair the soft electronic materials emerging from labs. Presentation slides are on Slideshare:
Jaz captured a video of his membrane pump moving water from left to right. The metallized membrane is a disc of about 1 cm diameter sandwiched between the pink layers of plastic, while metal tubes on either side supply an AC electric field. The voltage applied at the red and black clips is similar to what comes out of an electric socket, and the device draws a few milliamps. The video shows a single 10 micron thick membrane driving a slug of deionized water at about 60 microliters per minute. A microliter is 1 cubic millimeter, meaning this flow rate is in a good range for pushing fluids in “lab on a chip” analytical applications such as chromatography for identifying proteins. It could also be useful for driving tiny fluidic actuators, especially at hinges where a small volume change could drive a larger mechanical motion. And it would take a little over 8 minutes to adminster your .5 cc flu vaccine at this rate, which would be okay only if you were very patient. To get faster flow rates, you can increase the membrane surface area.
Our projects are about integrating “functional” materials into larger structures, without damaging the material or the structure. Some examples of functions we want the materials to have are exerting forces, absorbing specific wavelengths, or turning a mechanical stress into an electronic signal. That won’t happen if we melted the material during installation (a real danger when sealing gold nanoparticles into a polymer structure) or if the material destroys the structure that holds on to it. Brian needed to make sure his actuators could output useful forces without melting the plastic of a 3D printed robot body. Thanks to our neighbors at FirstBuild, he was able to check on the temperature of the actuators he made using their thermal imaging camera. After that, he went ahead and made sure the grad students were in good shape.