Designing for a Complex World
As an aspiring Human Factors Practitioner and Cognitive Systems Engineer, I have a passion for creating displays that effectively merge human and computer intelligence.
I’m currently working to expand my toolbox of analysis approaches and representation techniques, and I enjoy challenging myself with new projects to test and grow these skills.
The projects below are some of my most recently completed work, but I almost always have more on the way.
As part of a human-machine interface course in which I was assisting, we asked the students to re-design a computer interface that ICU nurses use to monitor and manage glucose levels in multiple patients. The task was presented to the students as a UX problem: They were asked to do a heuristic evaluation and task analysis, and then re-design the UI to make it easier for users to complete their most common tasks. However, as I learned about the work domain and critiqued the students’ designs, I started to notice that the problems with the existing interface go beyond usability. There are also underlying decision support needs that the current interface does not address.
On the sidelines of the class, I scribbled down the beginnings of a decision-centered redesign for this system, including a rough set of information requirements and some initial design concepts. I’ve recently decided to return to this project and finish it off. Specifically, I plan to perform a more complete cognitive work analysis, and then use the results of this analysis to develop detailed design concepts and static prototypes for a potential re-design of this system.
Check back here for more updates as this project moves ahead!
“Not everybody can make it as a server.” This comment, from a friend who is an experienced server and regularly trains new servers, started me down a fascinating deep dive into this work domain.
To explore the structure of this work domain and identify potential opportunities for decision support, I decided to perform a cognitive work analysis of restaurant serving, with my friend as a subject matter expert.
It turns out that functionally, restaurant serving has a lot in common with other work domains that we more often think of as “complex”, such as intelligence analysis. For example, this work involves using multiple data collection and inference loops to (1) identify the current resource demands of multiple competing processes and (2) allocate limited resources to those processes accordingly.
This analysis revealed some interesting opportunities for decision support in restaurant environments, specifically in the areas of data collection/inference and task sequencing.
Graduate students often balance many responsibilities in a limited amount of time. In this time-critical environment, inefficient meetings with advisors, fellow students, and project teams can be frustrating and significantly hinder the effectiveness of everyone involved. Research in organizational psychology has shown that properly using meeting agendas can help us to overcome this problem. However, despite these findings, many grad students don’t consistently use agendas when we meet with our advisors or group project teams.
To explore why agenda use can be infrequent among grad students and identify some ways to address this issue, I performed an analysis using some of the methods and principles that I picked up during a behavioral design course. Using an analysis workflow based on the Irrational Labs methodology, I modeled the process involved in creating and using meeting agendas, identified the psychological barriers that hinder each step of this process, and analyzed the emotional and functional benefits associated with agenda use. Based on this analysis, I identified several potential solutions to increase agenda use among graduate students.
Insurance underwriters need to assess the risks posed by potential policyholders to determine whether to write or renew their policies. These decisions are critical to an insurance fund’s success, but underwriters must often make them at a fast pace, with limited resources, and in a fragmented information landscape that places the burden of translating data into insight almost entirely on their shoulders. Without proper support from their tools, underwriters must lean heavily on their expertise and develop creative (but sub-optimal) workaround strategies to achieve an acceptable level of decision performance.
In Spring 2020, I worked with a team of students at UVA to design a decision support tool for the underwriters at a workers compensation insurance fund. Based on interviews with an expert underwriter and consultations with the client, we performed a cognitive work analysis to identify underwriters’ key information needs and designed an interface from the ground up to meet these requirements. The result was a clean, responsive system that that elegantly supports the cognitive work that underwriters must perform to successfully make mission-critical risk assessment and policy writing decisions.
In Spring 2020, I worked with a group of students at UVA to develop a way for prospective college students and their families to make informed decisions when choosing between schools and deciding to take on student loan debt. To address this need, we designed a desktop-based web app to be integrated into the existing Federal Student Aid website.
After an initial meeting with our client, we completed three iterations of analysis and design over the course of one month. We took a decision-centered approach, identifying the cognitive work users need to perform at each stage of their journey and designing our display elements to provide the information users need to complete that work. Our final deliverable was a state-based storyboard and presentation describing the core functionality of the app.
This was my first experience designing web applications for non-expert users, and it taught me several important lessons. These included the importance of adhering to useful display conventions, the need to simplify interfaces when designing for the general population, and the ability of effective visual design to augment the functionality of an app, among many others.
In Fall 2018, I worked with a team of Clemson University students to design a mobile phone interface for future artificial pancreas (AP) devices, which will automate blood glucose monitoring and insulin delivery for people with Type I diabetes. AP devices may not work properly under all conditions, and effective display interfaces will be needed to enable users to monitor and control the system. Our prototype display was submitted to and selected as a finalist in the 2019 Human Factors and Ergonomics in Healthcare Student Design Competition.
In Spring 2019, a meeting with a mentor in industry set me off on a one-man effort to redesign the interface using the principles of Ecological Interface Design, eventually leading to the prototype interface you see to the right. This project was my first exposure to CSE methodology, and it inspired me to pursue additional experience and training in this discipline.