We designed a system that helps people locate misplaced items at home called Doko. The system utilizes radio signal strength, triangulation, and sensors inside smartphones to calculate relative locations of tagged items. With this data, Doko can provide reference points of the item you are looking for and give you an intuitive instruction like "Your keys are in the living room close to the iPad charger, sketching book, and backpack".
This was a collaboration project with Tony Ip and Kosuke Shiraishi.
Primary & Secondary Research,
In order to get familiar with the technology and current market, we conducted several research activities including SME interviews, literature reviews, and competitive analysis.
We also conducted 5 field studies and follow-up interviews to collect data regarding tracking behavior in the context of home. For more details, please refer to our research summary.
We created a 2 x 2 framework that uses “Impact of Losing It” and “Frequency of Use” as axes, and put all items that people intend to track in our field studies onto this framework. We soon found that most of the items can be grouped into three main categories.
"I want to find it quickly" - These items such as TV remote, stationery, and portable mouse are used frequently and people want to find them within 1 minute. Finding it in a quickest way would be an important factor for this group.
"I need to know that it is safe at home" - The items in this category are rarely used but have a great impact on your life. You probably will not actively track it frequently, but you would want to get a notification when there is something wrong.
"I carry these things with me, and don’t want to lose it" - We think a lot of existing products such as Tile, stickNFind, and Chipolo are primarily focusing on this group, and we are trying to evaluating some of these products and find out how we can improve the user experience when tracking things.
We brainstormed and sketched 29 scenarios based on these categories. We then divided them into 9 groups according to the similarity and picked three promising groups that are fit for our interests and in the meantime can maximize the benefits of ambient backscatter technology.
With the idea of relative location, we believe that this is a better and intuitive way for people to find misplaced items at home. We refined our initial concepts, applied relative location into our final refinement, and sketched a storyboard and a scenario diagram.
After analyzing and comparing with the existing technologies, we found that in addition to the battery-free advantage, the most important feature of ambient backscatter is that the tags can communicate with each other and we can therefore draw a network map of the tagged items. In this way, we can know the relative locations of the items and group them according to proximity, and this is what existing product (Bluetooth Low Energy technology) can not achieve.
We started with an UI brainstorming session to figure out the flow of each scenario and components of each screen. We picked a lot of UI screens from several existing apps in various domains, and tried to extract and combine the components in order to get inspiration. We then sketched the draft screens and created an interactive wireframe for the first round evaluation.
Evaluation Round 1
We recruited 3 participants to conduct the first round evaluation. We set up a controlled environment where we asked participants to actually find items hid by us using our interactive wireframe and paper prototype.
Evaluation Round 2
We recruited 2 participants to conduct the second round evaluation. In order to refine our prototype, we tested multiple approaches to present the data of relative locations, reference points, and precise locations.