CASAS Smart Home Visualization Study!
Posted by A Leah Zulas, PhD on
4 minute read
CASAS Smart Home Visualization Study!
In this project we are working to create the best possible visualization of the health data gathered in the home. This will allow caregivers to see what is happening with the residents of the home.
We started with a visualization of occupancy data that was originally developed by Tiger Place in Missouri, which is an aging in place smart home affiliated with the University of Missouri. At the time, this visualization tool was limited to a black and white representation of whether a resident was active in the home at any given time. At Washington State University’s Center for Advanced Studies in Adaptive Systems (CASAS) the goal was to use a similar layout for the data, but to incorporate significantly more detail about the residents’ activities. The original visualization created at CASAS can be seen below (apartment identifiers are confidential tags and do not correspond to actual locations).
After user feedback, it was shown to be too difficult to understand this interface without in-depth study and/or training. It includes far too much information on a single page without sufficient indication of what each individual element denotes. With these deficiencies in mind, I began the process of improving the group’s tools for showing this complex data to users.
The first improvement I suggested to the team was to make this data easier to understand at a glance. After a concerted effort towards increasing legibility and translating a few of the graphs to a more intuitive gauge style, the result was a much more user-friendly interface, as seen below. This is not by any means a finished product, but it is a reasonable first step in improving the user experience of the visualization.
For our next iteration of the user interface design, we included the nursing staff who will be using this website in their function as caregivers. I conducted interviews with nine nursing staff from three locations in order to gather additional requirements for future development.
The interview consisted of five sections:
- General demographics
- Computer experience
- Ideal features the system should provide (gathered before the interviewee has been shown the current revision of the system)
- A short usability section to assess the quality of the current interface
- Future improvements for the system (after the interviewee has been shown the system)
When designing this interview process I attempted to keep it brief, but answer all vital research questions. The nursing staff is often very busy and lacks time for a lengthy interview; therefore the interview was kept to 30 minutes or less. Due to these constraints, we are attempting to measure computer experience with only three carefully crafted questions. Additional time savings were gained by using the System Usability Scale (SUS) assessment of only 10 questions. While both of these sections are brief, they are adequate for answering the current research questions on these topics.
With the requirements gathered via these staff interviews, the team and I worked toward a better product. This summer, progress consists of graphs with better axis and better overall clarity.
Once we finished assessing the needs of the nursing staff, we worked toward broadening our understanding of other caregivers by holding interviews with children and spouses of our elderly participants as well as geriatric physicians. With this information, we will have a more holistic view of our users. Another iteration of the web portal can be seen below along with copies of all of the relevant papers released from this series of studies.
I would like to thank Aaron Kwan and Tim McAleer for their help with the new design.
Papers on this work have been accepted as presentations at HFES2012: Nursing Assessment.
HFES2014: Child Caregiver Assessment.
CHI2016 (full paper, not abstract version submitted to WISH workshop): Spouse Caregiver Assessment.