 |
 |
 |
 |
 |
 |
Wardrobe-Aid: Closet Redesign
Although it is something we may take for granted, we make many complex decisions when we pick out our wardrobe. Often times, we regret making some decisions later in the day because of unforeseen changes in our schedule, or the weather, or obligations we have forgotten about. Some of us may also spend way too much time making these decisions. Do you or anyone you know spend an obscene amount of time picking out and reevaluating your clothing? Our group wanted to find a way to cater to these design processes so that time was saved and bad decisions were minimized.
Our team first lined up users (which consisted of 12 college-aged acquaintances). We then devised a series of questions and walkthroughs designed to find out how people make their wardrobe decisions in the morning, and how they carry out their dressing behavior.
We did contextual interviews of these users while choosing their outfit and allowed them to speak their mind as they traverse their dressing process. We used voice recorders and digital cameras to take pictures of the users' environments. The recorded interviews were quickly transcribed and shared with the entire group for processing. Our team then took the data, and began to analyze the behavior of the users along with the breakdowns of their processes. This analysis included task analysis, affinity diagrams, and various conceptual models including physical models, artifact models, and cultural models.
We ran into a lot of variety. People vary greatly in their living environments and even more in their thought processes about how to dress themselves. There is a lot of social influence that it not immediately obvious and may never actually be. Also, one of the main factors that affected a clothing choice was whether or not a certain piece was 'dirty'. The problem was people varied about what they consider dirty and what they consider clean. Further, for each person it also depended on the type of clothing.
We had to continually keep in mind that we were looking at behaviors and processes, and not the objects themselves. This made it more difficult than projects found in other groups (often, objects were the center of testing and design, such as trash cans or desks).
Also, we had to be very careful not to design for making decisions for the user, instead of aiding them in the decision-making process. For instance, we had all types of ideas to automate outfits based on a tagging system or previous choices. The problem is it's removing the user control from the whole process. Humans are very good at making these decisions and they are pulling all sorts of subtle pieces of cultural, and mood-shaped pieces of information into their heads. Trying to mimic this ability less-effectively while at the same time decreasing their control would result in a negative user experience.
Things that affect a person's wardrobe decision process:
- Weather (temperature, wind, precipitation)
- Culture
- Schedule
- What is clean/dirty
- What other friends are wearing (more specific than just cultural influences)
- emotional baseline (mood)
We also found that people almost universally use a flat surface when picking an outfit. Some people actually laid the clothes out correctly to represent how it would look on their body, and some did not. But they all found some type of flat surface to set their outfit on until they were ready to start putting it on. Most users used their bed- which caused a lot of needless back and forth activity.
In order to infer weather conditions, not only did people check reliable online sources, but they also looked out their window! Getting a view of the outside is a tremendously powerful, yet simple tool of inference.
- movement of trees = wind
- brightness of sun = heat
- color of clouds = precipitation
- color of sky = temperature, time of day
We designed a closet environment that catered to the users task of picking out their outfit. It is designed to be about the average-sized closet found in apartments, dorms, and houses that house college-aged persons. It includes a full-size mirror and hooks on the right door, and an interactive touch screen as well as a horizontal bar on the left. Within the closet is a set of visible drawers for organization of the user's choosing- as well as a set of two shelves that form a flush surface with the drawer top.
Our design is very low-tech since most people pick out their wardrobe in the morning and we all know that the last thing you'd like to do when you wake up is fidget with technology. The most hi-tech component if that of the interactive touch screen. It incorporates a camera feed of the outside (analogous to one's window), as well as other rich sets of information, like celebrity fashion outfit feeds, shared friend outfit feeds, and previous worn outfits. The user could also check his/her schedule and see the weather conditions for the rest of the day.
One feature that I would have liked to incorporate is the use of more ambient information, as opposed to deliberate screen interaction. For instance, instead of looking at the weather forecast to see if it's raining or how cold it is- why not close the semantic gap and make the conditions obvious inside the closet environment? You could incorporate quiet, but informative weather sounds: rain and thunder if it's actually raining (or going to rain). And because it's physically impossible to actually make the closet as cold or hot as the day would be in a reasonable time (and possibly not even desired), you could make the hangers of outfits glow if they've been worn in similar weather. So, if it's a particularly cold day, the closet would slightly suggest clothes that you've worn before on cold days. But the slightness of suggestion is imperative. It shouldn't seem intrusive to the user's decision-making process.
This project is a difficult one for many to see benefit because it is not directly projected towards a consumer basis. In other words, the final product as a whole is probably not something that you could go to a store and buy. It would be a commodity to future construction and architectural design firms that incorporate it into upper-end house models. These people usually have a hard time seeing value in usability and interaction design. However, as users get more demanding for better products and experiences, this will shift significantly (if it hasn't already!).