Automated Corset Shirt

 Our Automated Corset-style Shirt was an exploration into gesture based soft circuit design. It was intended to explore the gestures people generally make when they are cold. We found that when people are cold they tend to tighten up there arms, holding them close to the body.

Our design took this gesture and used soft-switches built into the shirt. These switches detected whether or not the wearer's arms are being held close to the body. To confirm that the user is indeed experiencing cold and not receiving outside pressure on their body or causing any other type of false-positive, we decided to use a temperature sensor also connected into the Arduino board. In order for the Arduino to decide the air temperature should be considered cold we exposed ourselves to different temperatures until we felt that we needed to either go to a warmer place or add a layer of clothing. We found that our average cool temperature was 15 Degrees Celcius. This temperature then became our lower bound for beginning to close the shirt.

In order for us to eliminate as much temperature cross-contamination from the wearer as possible, we placed the temperature sensor on the back side of the shoulder, as far from any major blood vessels as possible as well as being in the least restrictive place in regards to movement and interacting with others.

When the correct situation is detected the system calls upon two continuous rotation servo motors which are attached to a shoe lace running down the back of the shirt. The amount of rotation outputted to the servos was a ratio of the current temperature to our final closed temperature, 5 Degrees Celcius.
The trouble with using continuous servo motors is that they can have a problem finding their centre position which was the call that needs to be made in order to stop the rotation. In order to counter this we had to use a different method of rotation which does not work in degrees of rotation but in microseconds and sending electricity or not. Unfortunately by doing this method we sacrificed the ability to discreetly control rotation without using a conversion algorithm.

We ran into a few physical problems when installing the servos and soft-switches. With regards to the use of conductive thread we ran into difficulty as the natural resistance of the conductive thread reduced the amount of current that could be detected by the Arduino board as well as limited the already low amount of power that could be outputted to the servos. In a future iterations of this design the use of an external battery pack to boost power to the servos. The last two major problems had to do with the servos themselves. First, attaching servos of that size to the shirt proved to be beyond our scope at the time of fabrication. We later received the advice to build large pockets which can spread the weight of the servos over a larger surface area of the shirt. Lastly, once of the servos that we had used caused us problems in that it would not listen to the program and reverse its direction of rotation. This problem created a huge delay in work as much of our time went into troubleshooting the program and not the hardware, which was the actual cause of the problem. It turned out that this servo's potentiometer, which sets the 0 point of rotation for the servo, had been broken. This part of the servo being broken caused it to be unable to find its 0 point by which it had based its rotational direction.

A special reminder to anyone using equipment they are unfamiliar with: "No matter if it's new or used, always read the users manual. It will save you a lot of program troubleshooting and may give you some tricks to better work with the device.