We decided to participate in the CHI (Human Factors for Computing Systems) Student Design Competition, the theme of which was to create an intervention to “Level the Playing Field”. We found that the food insecure and homeless populations in the US, it is a daily struggle to get two square meals a day due to the absence of a information disemmination system. We decided to address the needs of this marginalized community.
Our research involved literature review of ICT solutions for the homeless, and interviews to understand the problems that these people faced in accessing the resources provided by the government, and reviewed literature. We identified some of the main challenges that this population faces. Our solution is a easy-to-access platform which will enable them access to information regarding resources for their immediate upliftment.
I led and conducted interviews, identified key areas for research, and served as the designer for this project.
We created a text-message based system for dumb-phones that increases access to hot meals for the homeless population. The system will send out information about the nearest food kitchens from a given zipcode along with other information needs.
Homeless populations are constantly on the move, and have various information needs ranging from transportation, education, housing, job search, and basic needs such as food. There are over half a million homeless people in the US and the most pervasive problem is that they don't know where to find free, hot meals because their information needs about emergency food are not being met.
We performed an extensive literature review to understand previous ICT-based solutions, the impact of and the reactions to such solutions, technology use amongst the homeless and other challenges that our target users face. Our primary findings were:
The homeless have various information needs ranging from transportation, education, housing, job search, and basic needs such as food.
Technology that is familiar and visually basic resulted in openness, and increased engagement with the system. In contrast, a visually attractive system created a perception of restriction.
Increase in requests for emergency food assistance in 71% of the cities, and in the no. of visits to emergency kitchens in 78% of the cities.
Smartphone ownership is rare; data plans are rarer. However, most people own mobile (dumb) phones through the "Obama phone" program.
Our findings from the literature review helped us to focus on solving the information needs regarding food. To gather further data and gain insights into the exact needs of the homeless population, we decided to conduct interviews.
We interviewed 13 people at a local community kitchen and shelter, and compensated each $5. Most of our interviewees mentioned that they frequently moved between cities. We asked them about their primary needs when in a new environment and what, if any, solutions do they currently use to find out about places which serve free food.
We also interviewed two social workers to get an insight into the needs expressed by the target audience to them.
The biggest challenge we faced when attempting to collect qualitative data from the homeless population was that they were usually unwilling to talk for more than 15 minutes, even when they were being paid. To overcome this, our initial 5 interviews were very exploratory, but we became much more focused and constrained with our questions for the latter interviews.
To meet their daily needs, homeless individuals are often on the move throughout the day.
They are very closeknit, and the desire to share information is high.
The local homeless community is usually aware of information about community kitchens, and other places which serve free hot meals; however, most of it is spread through word of mouth.
Most homeless people have access to mobile. All but one of our interviewees did not have any type of phone.
They often travel large distances because they are not aware of all the places which serve free, hot meals.
With our research findings, and understanding of user pain points and needs, we began our design phase. We first created personas to help us identify with the needs of our users. We then brainstormed several ideas as a group.
From the interview analysis we found two broad types of users emerge. We modeled them into two distinct personas which differed primarily in their outlook and motivations. These were: the people who were newly homeless; and the people who have been homeless for an extended duration.
Newly homeless people are more concerned about finding a job and do not think long term in terms of where their meals are coming from. People who have been homeless long-term have goals of ensuring constant access to free-food and shelter and try to find work within the vicinity.
We used a why/how laddering approach to generate different possible solutions that can help address this problem. Some of these were:
We settled on the solution which uses simple text messages to connect our users with information that is possessed among other members of the homeless community or social workers through text messages. As the homeless people have a strong network, we decided to leverage it in our solution.
We created sketches and interaction flows for our brainstormed solution. This enabled us to understand how our users will use the solution and how it will meet their needs.
The text-message system allows the user to enter in their zip code to receive a list of all locations serving free hot meals for that week. The information contains the name, address, days and times food is served, as well as identification requirements to receive a free meal.
The user can also choose to receive weekly notifications of all the locations and corresponding information for the zip code they entered..
The user can also update the database with new information they discover through their interactions with other people. This will be shared with other people in the area, however with the information that it is unverified.
If a user needs help using our system, they can text “Help” and receive instructions on what to enter based on different options available to interact with the system.
We evaluated our designed solution with 4 homeless people. We used the wizard of oz technique to test the system.
We provided our participants with a brief introduction to the system, how to interact with it, and then presented them with different scenarios. We then asked them to use the system and to complete the tasks, and intervened only when they struggled.
We found that they appreciated the ability to use the system from any device including flip-phones. We noticed that several people were semi-literate and struggled with interpreting the instructions. We believe that some degree of training might be required.
We hope that our system can be expanded to meet the broader information needs of our target users.
Although there are definitely some shortcomings to our solution, we believe that under the constraints of our target users, we conceptualized a solution which would have the widest reach.