A project for a kiosk-based conversational UI for human services needs which included research, concept design, and prototype/mockup. Created in order to address the homeless and elderly population in urban areas such as San Francisco, and bringing more accessibility to human services information with a multi-modal approach (screen user interface plus voice user interface). This was part of a Master's Degree course at Iowa State University (in the College of Engineering), and the project was co-created with Jessica Boddicker. Each team member participated in all activities.
Objective: How do we make specialized human services (ie. housing, food, health services) more accessible for people in-need, within San Francisco?
Challenges: research was conducted in order to better understand obstacles with being homeless, transient, impoverished or elderly and living in San Francisco, California. This research including trying to understand the systemic factors that may make it difficult or even prevent these population to get the help they may need.
We wanted to know, how do people can get help with specific needs (like a place to sleep, food to eat, a shower, etc), if they have limited access to the internet? How can we use research and design, in order to provide better human services to those in-need?
Process: The general design process for this project started with: 1. addressing the content through market and user research, to narrow down product requirements and understand the market, 2. creating wireframes and sketches, 3. iterations of the wireframes, then creating prototypes and mock ups as a result, 4. obtain feedback and analytics to further iterations on the design.
We reviewed data, including the 2017 SF Homeless Census, in order for persona creation. Three personas were created, one of which is a secondary user type. This projects contributions to micro and macro ergonomics were outlined early in the project and include improving local infrastructure and a focus on accessibility & efficiency. A mind map diagram was created in order to organize and visualize tasks based on what our personas need and want. Because we are creating a product in order to facilitate discovery of services, the taxonomy and information architecture (IA) were a major part of our project (for both the conversational design and the visual design) and were very considered in the wireframe and sketching stage. We wanted this kiosk to have ease of use, quick use, simple language, a limited visual user interface, and clarity of labels. Many of our design revisions were a process of simplification.
Outcome: by learning more about this topic, we decided which tools would be most beneficial to use for design while considering user needs, human factors and accessibility issues. This was determined to be a conversational interface (voice user interface or VUI) kiosk.
Some of the feedback from user testing which help guide the final design included: simplifying/reducing secondary text on screen, simpler language usage, taxonomy and labelling need to logically map to content, kiosk homepage needed to be more scannable and feature larger text, the visualization of voice interaction needed to be removed because it didn't feel relevant/too busy and in order to encourage visual scanning we needed to reduce button salience by shifting it down on the page.
We created a prototype of our kiosk visual screen user interface as well as a conversational interface prototype and tested them, in order to improve and revise our designs further based on the research outcomes. Both the visual user interface and conversational interface were both greatly modified throughout the process, based on the user feedback and analytics.
Final deliverables available upon request, included: final research paper (70 pages) and contained personas, mind map, information architecture diagram, hierarchical task analysis (HTA) of both the UI and VUI, interface design (with design revisions) for the UI and VUI, prototypes of both the touchscreen UI and conversational interface, user testing results, statistical analysis, and two product design mock ups of the kiosk.
Date: 2018 • Iowa State University's College of Engineering educational project co- produced by Jessica Boddicker
Tools: JASP (for statistical analysis), Sketch, OmniGraffle, Google Drive, InVision